wireshark/epan/column-utils.c

2109 lines
56 KiB
C
Raw Normal View History

/* column-utils.c
* Routines for column utilities.
*
* Wireshark - Network traffic analyzer
* By Gerald Combs <gerald@wireshark.org>
* Copyright 1998 Gerald Combs
*
* SPDX-License-Identifier: GPL-2.0-or-later
*/
#include "config.h"
#include <stdio.h>
#include <string.h>
#include <time.h>
#include <locale.h>
#include <limits.h>
#include "column-utils.h"
#include "timestamp.h"
#include "to_str.h"
#include "packet_info.h"
#include "wsutil/pint.h"
#include "addr_resolv.h"
#include "address_types.h"
#include "osi-utils.h"
#include "value_string.h"
#include "column-info.h"
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
#include "column.h"
#include "proto.h"
#include <epan/strutil.h>
#include <epan/epan.h>
#include <epan/dfilter/dfilter.h>
#include <wsutil/inet_cidr.h>
#include <wsutil/utf8_entities.h>
2021-06-18 18:21:42 +00:00
#include <wsutil/ws_assert.h>
#include <wsutil/unicode-utils.h>
#include <wsutil/time_util.h>
#ifdef HAVE_LUA
#include <epan/wslua/wslua.h>
#endif
#define COL_BUF_MAX_LEN (((COL_MAX_INFO_LEN) > (COL_MAX_LEN)) ? \
(COL_MAX_INFO_LEN) : (COL_MAX_LEN))
/* Used for locale decimal point */
static char *col_decimal_point;
/* Used to indicate updated column information, e.g. a new request/response. */
static gboolean col_data_changed_;
static int proto_cols;
static gint ett_cols;
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
/* Allocate all the data structures for constructing column data, given
the number of columns. */
void
col_setup(column_info *cinfo, const gint num_cols)
{
int i;
col_decimal_point = localeconv()->decimal_point;
cinfo->num_cols = num_cols;
cinfo->columns = g_new(col_item_t, num_cols);
cinfo->col_first = g_new(int, NUM_COL_FMTS);
cinfo->col_last = g_new(int, NUM_COL_FMTS);
for (i = 0; i < num_cols; i++) {
cinfo->columns[i].col_custom_fields_ids = NULL;
}
cinfo->col_expr.col_expr = g_new(const gchar*, num_cols + 1);
cinfo->col_expr.col_expr_val = g_new(gchar*, num_cols + 1);
for (i = 0; i < NUM_COL_FMTS; i++) {
cinfo->col_first[i] = -1;
cinfo->col_last[i] = -1;
}
cinfo->prime_regex = g_regex_new(COL_CUSTOM_PRIME_REGEX,
(GRegexCompileFlags) (G_REGEX_RAW),
0, NULL);
}
columns: Allow any field expression syntax to be used in columns Allow anything that can be used in a display filter to be used in columns (with the exception that field references don't work without a notion of a currently selected frame): display filter functions, slices, arithmetic calculations, logical tests, raw byte addressing, the layer modifier, display filter macros, etc., alone or in combination. Show the results and generate filters. Note that "resolved" values are not yet supported. They make conceptual sense for some expressions (e.g., if the layer modifier only is used) but not for others. Perhaps resolution could be done as a final step in the filter before returning values. It would also be useful to be able to get the expected return type of an expression, so that the functions for right justifying a column or sorting numerically could work. Right now the results are treated as strings even if the return field values are numeric. Multifield columns (i.e., concatenation of field values) are currently implemented using the OR operator.For backwards compability, continue to support that. When a true logical OR would give a different result, surround the expression in parentheses, which the multifield columns did not previously support (due to the regex used instead of full filter grammar parsing.) Perhaps in the future we should introduce a separate operator for concatenation, possibly only used in column definitions and nowhere else. Update release notes. Fix #7752. Fix #10154. Fix #15990. Fix #18588. Fix #19076. Related to #16181 - it's now possibly to define new display filter functions so that is essentially solved, though I suppose there's always room for more built-in functions.
2024-02-08 02:13:23 +00:00
static void
col_custom_free_cb(void *data)
{
col_custom_t *col_custom = (col_custom_t*)data;
dfilter_free(col_custom->dfilter);
g_free(col_custom->dftext);
g_free(col_custom);
}
static void
col_custom_fields_ids_free(GSList** custom_fields_id)
{
if (*custom_fields_id != NULL) {
columns: Allow any field expression syntax to be used in columns Allow anything that can be used in a display filter to be used in columns (with the exception that field references don't work without a notion of a currently selected frame): display filter functions, slices, arithmetic calculations, logical tests, raw byte addressing, the layer modifier, display filter macros, etc., alone or in combination. Show the results and generate filters. Note that "resolved" values are not yet supported. They make conceptual sense for some expressions (e.g., if the layer modifier only is used) but not for others. Perhaps resolution could be done as a final step in the filter before returning values. It would also be useful to be able to get the expected return type of an expression, so that the functions for right justifying a column or sorting numerically could work. Right now the results are treated as strings even if the return field values are numeric. Multifield columns (i.e., concatenation of field values) are currently implemented using the OR operator.For backwards compability, continue to support that. When a true logical OR would give a different result, surround the expression in parentheses, which the multifield columns did not previously support (due to the regex used instead of full filter grammar parsing.) Perhaps in the future we should introduce a separate operator for concatenation, possibly only used in column definitions and nowhere else. Update release notes. Fix #7752. Fix #10154. Fix #15990. Fix #18588. Fix #19076. Related to #16181 - it's now possibly to define new display filter functions so that is essentially solved, though I suppose there's always room for more built-in functions.
2024-02-08 02:13:23 +00:00
g_slist_free_full(*custom_fields_id, col_custom_free_cb);
}
*custom_fields_id = NULL;
}
/* Cleanup all the data structures for constructing column data; undoes
the allocations that col_setup() does. */
void
col_cleanup(column_info *cinfo)
{
int i;
col_item_t* col_item;
if (!cinfo)
return;
for (i = 0; i < cinfo->num_cols; i++) {
col_item = &cinfo->columns[i];
g_free(col_item->fmt_matx);
g_free(col_item->col_title);
g_free(col_item->col_custom_fields);
dfilter_free(col_item->col_custom_dfilter);
/* col_item->col_data points to col_buf or static memory */
g_free(col_item->col_buf);
g_free(cinfo->col_expr.col_expr_val[i]);
col_custom_fields_ids_free(&col_item->col_custom_fields_ids);
}
g_free(cinfo->columns);
g_free(cinfo->col_first);
g_free(cinfo->col_last);
/*
* XXX - MSVC doesn't correctly handle the "const" qualifier; it thinks
* "const XXX **" means "pointer to const pointer to XXX", i.e. that
* it's a pointer to something that's "const"ant, not "pointer to
* pointer to const XXX", i.e. that it's a pointer to a pointer to
* something that's "const"ant. Cast its bogus complaints away.
*/
g_free((gchar **)cinfo->col_expr.col_expr);
g_free(cinfo->col_expr.col_expr_val);
if (cinfo->prime_regex)
g_regex_unref(cinfo->prime_regex);
}
/* Initialize the data structures for constructing column data. */
void
col_init(column_info *cinfo, const struct epan_session *epan)
{
int i;
col_item_t* col_item;
if (!cinfo)
return;
for (i = 0; i < cinfo->num_cols; i++) {
col_item = &cinfo->columns[i];
col_item->col_buf[0] = '\0';
col_item->col_data = col_item->col_buf;
col_item->col_fence = 0;
col_item->writable = TRUE;
cinfo->col_expr.col_expr[i] = "";
cinfo->col_expr.col_expr_val[i][0] = '\0';
}
cinfo->writable = TRUE;
cinfo->epan = epan;
}
gboolean
col_get_writable(column_info *cinfo, const gint col)
{
int i;
col_item_t* col_item;
if (cinfo == NULL)
return FALSE;
/* "global" (not) writeability will always override
an individual column */
if ((col == -1) || (cinfo->writable == FALSE))
return cinfo->writable;
if (cinfo->col_first[col] >= 0) {
for (i = cinfo->col_first[col]; i <= cinfo->col_last[col]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[col]) {
return col_item->writable;
}
}
}
return FALSE;
}
void
col_set_writable(column_info *cinfo, const gint col, const gboolean writable)
{
int i;
col_item_t* col_item;
if (cinfo) {
if (col == -1) {
cinfo->writable = writable;
} else if (cinfo->col_first[col] >= 0) {
for (i = cinfo->col_first[col]; i <= cinfo->col_last[col]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[col]) {
col_item->writable = writable;
}
}
}
}
}
/* Checks to see if a particular packet information element is needed for the packet list */
#define CHECK_COL(cinfo, el) \
/* We are constructing columns, and they're writable */ \
(col_get_writable(cinfo, el) && \
/* There is at least one column in that format */ \
((cinfo)->col_first[el] >= 0))
/* Sets the fence for a column to be at the end of the column. */
void
col_set_fence(column_info *cinfo, const gint el)
{
int i;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
col_item->col_fence = (int)strlen(col_item->col_data);
}
}
}
/* Clear the fence for a column. */
void
col_clear_fence(column_info *cinfo, const gint el)
{
int i;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
col_item->col_fence = 0;
}
}
}
/* Gets the text of a column */
const gchar *
col_get_text(column_info *cinfo, const gint el)
{
int i;
const gchar* text = NULL;
col_item_t* col_item;
if (!(cinfo && (cinfo)->col_first[el] >= 0)) {
return NULL;
}
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
text = (col_item->col_data);
}
}
return text;
}
/* Use this to clear out a column, especially if you're going to be
appending to it later; at least on some platforms, it's more
efficient than using "col_add_str()" with a null string, and
more efficient than "col_set_str()" with a null string if you
later append to it, as the later append will cause a string
copy to be done. */
void
col_clear(column_info *cinfo, const gint el)
{
int i;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
/*
* At this point, either
*
* 1) col_data[i] is equal to col_buf[i], in which case we
* don't have to worry about copying col_data[i] to
* col_buf[i];
*
* 2) col_data[i] isn't equal to col_buf[i], in which case
* the only thing that's been done to the column is
* "col_set_str()" calls and possibly "col_set_fence()"
* calls, in which case the fence is either unset and
* at the beginning of the string or set and at the end
* of the string - if it's at the beginning, we're just
* going to clear the column, and if it's at the end,
* we don't do anything.
*/
if (col_item->col_buf == col_item->col_data || col_item->col_fence == 0) {
/*
* The fence isn't at the end of the column, or the column wasn't
* last set with "col_set_str()", so clear the column out.
*/
col_item->col_buf[col_item->col_fence] = '\0';
col_item->col_data = col_item->col_buf;
}
cinfo->col_expr.col_expr[i] = "";
cinfo->col_expr.col_expr_val[i][0] = '\0';
}
}
}
#define COL_CHECK_APPEND(col_item, max_len) \
if (col_item->col_data != col_item->col_buf) { \
/* This was set with "col_set_str()"; copy the string they \
set it to into the buffer, so we can append to it. */ \
(void) g_strlcpy(col_item->col_buf, col_item->col_data, max_len); \
col_item->col_data = col_item->col_buf; \
}
#define COL_CHECK_REF_TIME(fd, buf) \
if (fd->ref_time) { \
(void) g_strlcpy(buf, "*REF*", COL_MAX_LEN ); \
return; \
}
/* The same as CHECK_COL(), but without the check to see if the column is writable. */
#define HAVE_CUSTOM_COLS(cinfo) ((cinfo) && (cinfo)->col_first[COL_CUSTOM] >= 0)
gboolean
have_custom_cols(column_info *cinfo)
{
return HAVE_CUSTOM_COLS(cinfo);
}
gboolean
have_field_extractors(void)
{
#ifdef HAVE_LUA
return wslua_has_field_extractors();
#else
return FALSE;
#endif
}
/* search in edt tree custom fields */
void col_custom_set_edt(epan_dissect_t *edt, column_info *cinfo)
{
int i;
col_item_t* col_item;
if (!HAVE_CUSTOM_COLS(cinfo))
return;
for (i = cinfo->col_first[COL_CUSTOM];
i <= cinfo->col_last[COL_CUSTOM]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[COL_CUSTOM] &&
col_item->col_custom_fields &&
col_item->col_custom_fields_ids) {
col_item->col_data = col_item->col_buf;
cinfo->col_expr.col_expr[i] = epan_custom_set(edt, col_item->col_custom_fields_ids,
col_item->col_custom_occurrence,
col_item->col_buf,
cinfo->col_expr.col_expr_val[i],
COL_MAX_LEN);
}
}
}
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
#if 0
// Needed if we create _ws.col.custom
static void
col_custom_set(proto_tree *tree, column_info *cinfo)
{
int i;
col_item_t* col_item;
if (!HAVE_CUSTOM_COLS(cinfo))
return;
for (i = cinfo->col_first[COL_CUSTOM];
i <= cinfo->col_last[COL_CUSTOM]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[COL_CUSTOM] &&
col_item->col_custom_fields &&
col_item->col_custom_fields_ids) {
col_item->col_data = col_item->col_buf;
cinfo->col_expr.col_expr[i] = proto_custom_set(tree, col_item->col_custom_fields_ids,
col_item->col_custom_occurrence,
col_item->col_buf,
cinfo->col_expr.col_expr_val[i],
COL_MAX_LEN);
}
}
}
#endif
void
col_custom_prime_edt(epan_dissect_t *edt, column_info *cinfo)
{
int i;
col_item_t* col_item;
if (!HAVE_CUSTOM_COLS(cinfo))
return;
for (i = cinfo->col_first[COL_CUSTOM];
i <= cinfo->col_last[COL_CUSTOM]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[COL_CUSTOM] &&
col_item->col_custom_dfilter) {
epan_dissect_prime_with_dfilter(edt, col_item->col_custom_dfilter);
}
}
}
char*
col_custom_get_filter(epan_dissect_t *edt, column_info *cinfo, const gint col)
{
col_item_t* col_item;
ws_assert(cinfo);
ws_assert(col < cinfo->num_cols);
col_item = &cinfo->columns[col];
if (col_item->fmt_matx[COL_CUSTOM] &&
col_item->col_custom_fields &&
col_item->col_custom_fields_ids) {
return proto_custom_get_filter(edt, col_item->col_custom_fields_ids,
col_item->col_custom_occurrence);
}
return NULL;
}
void
col_append_lstr(column_info *cinfo, const gint el, const gchar *str1, ...)
{
va_list ap;
size_t pos, max_len;
int i;
const gchar *str;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
/*
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
pos = strlen(col_item->col_buf);
if (pos >= max_len)
return;
va_start(ap, str1);
str = str1;
do {
if (G_UNLIKELY(str == NULL)) {
str = "(null)";
}
WS_UTF_8_CHECK(str, -1);
pos = ws_label_strcpy(col_item->col_buf, max_len, pos, str, 0);
} while (pos < max_len && (str = va_arg(ap, const char *)) != COL_ADD_LSTR_TERMINATOR);
va_end(ap);
}
}
}
void
col_append_str_uint(column_info *cinfo, const gint col, const gchar *abbrev, guint32 val, const gchar *sep)
{
char buf[16];
guint32_to_str_buf(val, buf, sizeof(buf));
col_append_lstr(cinfo, col, sep ? sep : "", abbrev, "=", buf, COL_ADD_LSTR_TERMINATOR);
}
static inline void
col_snprint_port(gchar *buf, size_t buf_siz, port_type typ, guint16 val)
{
const char *str;
if (gbl_resolv_flags.transport_name &&
(str = try_serv_name_lookup(typ, val)) != NULL) {
snprintf(buf, buf_siz, "%s(%"PRIu16")", str, val);
} else {
snprintf(buf, buf_siz, "%"PRIu16, val);
}
}
void
col_append_ports(column_info *cinfo, const gint col, port_type typ, guint16 src, guint16 dst)
{
char buf_src[32], buf_dst[32];
col_snprint_port(buf_src, 32, typ, src);
col_snprint_port(buf_dst, 32, typ, dst);
col_append_lstr(cinfo, col, buf_src, " " UTF8_RIGHTWARDS_ARROW " ", buf_dst, COL_ADD_LSTR_TERMINATOR);
}
void
col_append_frame_number(packet_info *pinfo, const gint col, const gchar *fmt_str, guint frame_num)
{
col_append_fstr(pinfo->cinfo, col, fmt_str, frame_num);
if (!pinfo->fd->visited) {
col_data_changed_ = TRUE;
}
}
static void
col_do_append_fstr(column_info *cinfo, const int el, const char *separator, const char *format, va_list ap)
{
size_t len, max_len, sep_len, pos;
int i;
col_item_t* col_item;
char tmp[COL_BUF_MAX_LEN];
sep_len = (separator) ? strlen(separator) : 0;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
/*
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
len = strlen(col_item->col_buf);
/*
* If we have a separator, append it if the column isn't empty.
*/
if (sep_len != 0 && len != 0) {
(void) ws_label_strcat(col_item->col_buf, max_len, separator, 0);
len += sep_len;
}
if (len < max_len) {
va_list ap2;
2021-12-12 11:34:48 +00:00
va_copy(ap2, ap);
pos = vsnprintf(tmp, sizeof(tmp), format, ap2);
va_end(ap2);
if (pos >= max_len) {
ws_utf8_truncate(tmp, max_len - 1);
}
WS_UTF_8_CHECK(tmp, -1);
ws_label_strcpy(col_item->col_buf, max_len, len, tmp, 0);
}
}
}
}
/* Appends a vararg list to a packet info string. */
void
col_append_fstr(column_info *cinfo, const gint el, const gchar *format, ...)
{
va_list ap;
if (!CHECK_COL(cinfo, el))
return;
va_start(ap, format);
col_do_append_fstr(cinfo, el, NULL, format, ap);
va_end(ap);
}
/* Appends a vararg list to a packet info string.
* Prefixes it with the given separator if the column is not empty.
*/
void
col_append_sep_fstr(column_info *cinfo, const gint el, const gchar *separator,
const gchar *format, ...)
{
va_list ap;
if (!CHECK_COL(cinfo, el))
return;
if (separator == NULL)
separator = ", "; /* default */
va_start(ap, format);
col_do_append_fstr(cinfo, el, separator, format, ap);
va_end(ap);
}
/* Prepends a vararg list to a packet info string. */
void
col_prepend_fstr(column_info *cinfo, const gint el, const gchar *format, ...)
{
va_list ap;
int i;
char orig_buf[COL_BUF_MAX_LEN];
const char *orig;
size_t max_len, pos;
col_item_t* col_item;
char tmp[COL_BUF_MAX_LEN];
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
if (col_item->col_data != col_item->col_buf) {
/* This was set with "col_set_str()"; which is effectively const */
orig = col_item->col_data;
} else {
(void) g_strlcpy(orig_buf, col_item->col_buf, max_len);
orig = orig_buf;
}
va_start(ap, format);
pos = vsnprintf(tmp, sizeof(tmp), format, ap);
va_end(ap);
if (pos >= max_len) {
ws_utf8_truncate(tmp, max_len - 1);
}
WS_UTF_8_CHECK(tmp, -1);
pos = ws_label_strcpy(col_item->col_buf, max_len, 0, tmp, 0);
/*
* Move the fence, unless it's at the beginning of the string.
*/
if (col_item->col_fence > 0)
col_item->col_fence += (int) strlen(col_item->col_buf);
/*
* Append the original data.
*/
ws_label_strcpy(col_item->col_buf, max_len, pos, orig, 0);
col_item->col_data = col_item->col_buf;
}
}
}
void
col_prepend_fence_fstr(column_info *cinfo, const gint el, const gchar *format, ...)
{
va_list ap;
int i;
char orig_buf[COL_BUF_MAX_LEN];
const char *orig;
size_t max_len, pos;
col_item_t* col_item;
char tmp[COL_BUF_MAX_LEN];
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
if (col_item->col_data != col_item->col_buf) {
/* This was set with "col_set_str()"; which is effectively const */
orig = col_item->col_data;
} else {
(void) g_strlcpy(orig_buf, col_item->col_buf, max_len);
orig = orig_buf;
}
va_start(ap, format);
pos = vsnprintf(tmp, sizeof(tmp), format, ap);
va_end(ap);
if (pos >= max_len) {
ws_utf8_truncate(tmp, max_len - 1);
}
WS_UTF_8_CHECK(tmp, -1);
pos = ws_label_strcpy(col_item->col_buf, max_len, 0, tmp, 0);
/*
* Move the fence if it exists, else create a new fence at the
* end of the prepended data.
*/
if (col_item->col_fence > 0) {
col_item->col_fence += (int) strlen(col_item->col_buf);
} else {
col_item->col_fence = (int) strlen(col_item->col_buf);
}
/*
* Append the original data.
*/
ws_label_strcpy(col_item->col_buf, max_len, pos, orig, 0);
col_item->col_data = col_item->col_buf;
}
}
}
/* Use this if "str" points to something that won't stay around (and
must thus be copied). */
void
col_add_str(column_info *cinfo, const gint el, const gchar* str)
{
int i;
size_t max_len;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
if (col_item->col_fence != 0) {
/*
* We will append the string after the fence.
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
} else {
/*
* There's no fence, so we can just write to the string.
*/
col_item->col_data = col_item->col_buf;
}
WS_UTF_8_CHECK(str, -1);
(void) ws_label_strcpy(col_item->col_buf, max_len, col_item->col_fence, str, 0);
}
}
}
/* Use this if "str" points to something that will stay around (and thus
needn't be copied). */
void
col_set_str(column_info *cinfo, const gint el, const gchar* str)
{
int i;
size_t max_len;
col_item_t* col_item;
DISSECTOR_ASSERT(str);
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
if (col_item->col_fence != 0) {
/*
* We will append the string after the fence.
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
(void) g_strlcpy(&col_item->col_buf[col_item->col_fence], str, max_len - col_item->col_fence);
} else {
/*
* There's no fence, so we can just set the column to point
* to the string.
*/
col_item->col_data = str;
}
}
}
}
void
col_add_lstr(column_info *cinfo, const gint el, const gchar *str1, ...)
{
va_list ap;
int i;
gsize pos;
gsize max_len;
const gchar *str;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
pos = col_item->col_fence;
if (pos != 0) {
/*
* We will append the string after the fence.
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
} else {
/*
* There's no fence, so we can just write to the string.
*/
col_item->col_data = col_item->col_buf;
}
va_start(ap, str1);
str = str1;
do {
if (G_UNLIKELY(str == NULL)) {
str = "(null)";
}
WS_UTF_8_CHECK(str, -1);
pos = ws_label_strcpy(col_item->col_buf, max_len, pos, str, 0);
} while (pos < max_len && (str = va_arg(ap, const char *)) != COL_ADD_LSTR_TERMINATOR);
va_end(ap);
}
}
}
/* Adds a vararg list to a packet info string. */
void
col_add_fstr(column_info *cinfo, const gint el, const gchar *format, ...)
{
va_list ap;
int i, pos;
int max_len;
col_item_t* col_item;
char tmp[COL_BUF_MAX_LEN];
if (!CHECK_COL(cinfo, el))
return;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
if (col_item->col_fence != 0) {
/*
* We will append the string after the fence.
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
} else {
/*
* There's no fence, so we can just write to the string.
*/
col_item->col_data = col_item->col_buf;
}
va_start(ap, format);
pos = vsnprintf(tmp, sizeof(tmp), format, ap);
va_end(ap);
if (pos >= max_len) {
ws_utf8_truncate(tmp, max_len - 1);
}
WS_UTF_8_CHECK(tmp, -1);
ws_label_strcpy(col_item->col_buf, max_len, col_item->col_fence, tmp, 0);
}
}
}
static void
col_do_append_str(column_info *cinfo, const gint el, const gchar* separator,
const gchar* str)
{
int i;
size_t len, max_len;
col_item_t* col_item;
if (el == COL_INFO)
max_len = COL_MAX_INFO_LEN;
else
max_len = COL_MAX_LEN;
for (i = cinfo->col_first[el]; i <= cinfo->col_last[el]; i++) {
col_item = &cinfo->columns[i];
if (col_item->fmt_matx[el]) {
/*
* First arrange that we can append, if necessary.
*/
COL_CHECK_APPEND(col_item, max_len);
len = col_item->col_buf[0];
/*
* If we have a separator, append it if the column isn't empty.
*/
if (separator != NULL) {
if (len != 0) {
(void) ws_label_strcat(col_item->col_buf, max_len, separator, 0);
}
}
WS_UTF_8_CHECK(str, -1);
(void) ws_label_strcat(col_item->col_buf, max_len, str, 0);
}
}
}
void
col_append_str(column_info *cinfo, const gint el, const gchar* str)
{
if (!CHECK_COL(cinfo, el))
return;
col_do_append_str(cinfo, el, NULL, str);
}
void
col_append_sep_str(column_info *cinfo, const gint el, const gchar* separator,
const gchar* str)
{
if (!CHECK_COL(cinfo, el))
return;
if (separator == NULL)
separator = ", "; /* default */
col_do_append_str(cinfo, el, separator, str);
}
/* --------------------------------- */
gboolean
col_has_time_fmt(column_info *cinfo, const gint col)
{
col_item_t* col_item = &cinfo->columns[col];
return ((col_item->fmt_matx[COL_CLS_TIME]) ||
(col_item->fmt_matx[COL_ABS_TIME]) ||
(col_item->fmt_matx[COL_ABS_YMD_TIME]) ||
(col_item->fmt_matx[COL_ABS_YDOY_TIME]) ||
(col_item->fmt_matx[COL_UTC_TIME]) ||
(col_item->fmt_matx[COL_UTC_YMD_TIME]) ||
(col_item->fmt_matx[COL_UTC_YDOY_TIME]) ||
(col_item->fmt_matx[COL_REL_TIME]) ||
(col_item->fmt_matx[COL_DELTA_TIME]) ||
(col_item->fmt_matx[COL_DELTA_TIME_DIS]));
}
static int
get_frame_timestamp_precision(const frame_data *fd)
{
int tsprecision;
tsprecision = timestamp_get_precision();
if (tsprecision == TS_PREC_AUTO)
tsprecision = fd->tsprec;
else if (tsprecision < 0)
ws_assert_not_reached();
/*
* Time stamp precision values higher than the maximum
* precision we support can't be handled. Just display
* those times with the maximum precision we support.
*/
if (tsprecision > WS_TSPREC_MAX)
tsprecision = WS_TSPREC_MAX;
return tsprecision;
}
static int
get_default_timestamp_precision(void)
{
int tsprecision;
tsprecision = timestamp_get_precision();
if (tsprecision == TS_PREC_AUTO)
tsprecision = WS_TSPREC_MAX; /* default to the maximum precision we support */
else if (tsprecision < 0)
ws_assert_not_reached();
/*
* Time stamp precision values higher than the maximum
* precision we support can't be handled. Just display
* those times with the maximum precision we support.
*/
if (tsprecision > WS_TSPREC_MAX)
tsprecision = WS_TSPREC_MAX;
return tsprecision;
}
static void
set_abs_ymd_time(const frame_data *fd, gchar *buf, char *decimal_point, gboolean local)
{
if (!fd->has_ts) {
buf[0] = '\0';
return;
}
format_nstime_as_iso8601(buf, COL_MAX_LEN, &fd->abs_ts, decimal_point, local, get_frame_timestamp_precision(fd));
}
static void
col_set_abs_ymd_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_ymd_time(fd, cinfo->columns[col].col_buf, col_decimal_point, TRUE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
col_set_utc_ymd_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_ymd_time(fd, cinfo->columns[col].col_buf, col_decimal_point, FALSE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
set_abs_ydoy_time(const frame_data *fd, gchar *buf, char *decimal_point, gboolean local)
{
struct tm tm, *tmp;
char *ptr;
size_t remaining;
int num_bytes;
int tsprecision;
if (!fd->has_ts) {
buf[0] = '\0';
return;
}
if (local)
tmp = ws_localtime_r(&fd->abs_ts.secs, &tm);
else
tmp = ws_gmtime_r(&fd->abs_ts.secs, &tm);
if (tmp == NULL) {
snprintf(buf, COL_MAX_LEN, "Not representable");
return;
}
ptr = buf;
remaining = COL_MAX_LEN;
num_bytes = snprintf(ptr, remaining,"%04d/%03d %02d:%02d:%02d",
tmp->tm_year + 1900,
tmp->tm_yday + 1,
tmp->tm_hour,
tmp->tm_min,
tmp->tm_sec);
if (num_bytes < 0) {
/*
* That got an error.
* Not much else we can do.
*/
snprintf(ptr, remaining, "snprintf() failed");
return;
}
if ((unsigned int)num_bytes >= remaining) {
/*
* That filled up or would have overflowed the buffer.
* Nothing more we can do.
*/
return;
}
ptr += num_bytes;
remaining -= num_bytes;
tsprecision = get_frame_timestamp_precision(fd);
if (tsprecision != 0) {
/*
* Append the fractional part.
* Get the nsecs as a 32-bit unsigned value, as it should never
* be negative, so we treat it as unsigned.
*/
format_fractional_part_nsecs(ptr, remaining, (guint32)fd->abs_ts.nsecs, decimal_point, tsprecision);
}
}
static void
col_set_abs_ydoy_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_ydoy_time(fd, cinfo->columns[col].col_buf, col_decimal_point, TRUE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
col_set_utc_ydoy_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_ydoy_time(fd, cinfo->columns[col].col_buf, col_decimal_point, FALSE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
set_time_seconds(const frame_data *fd, const nstime_t *ts, gchar *buf)
{
ws_assert(fd->has_ts);
display_signed_time(buf, COL_MAX_LEN, ts, get_frame_timestamp_precision(fd));
}
static void
set_time_hour_min_sec(const frame_data *fd, const nstime_t *ts, gchar *buf, char *decimal_point)
{
time_t secs = ts->secs;
guint32 nsecs;
gboolean negative = FALSE;
char *ptr;
size_t remaining;
int num_bytes;
int tsprecision;
ws_assert(fd->has_ts);
if (secs < 0) {
secs = -secs;
negative = TRUE;
}
if (ts->nsecs >= 0) {
nsecs = ts->nsecs;
} else if (G_LIKELY(ts->nsecs != INT_MIN)) {
/*
* This isn't the smallest negative number that fits in 32
* bits, so we can compute its negative and store it in a
* 32-bit unsigned int variable.
*/
nsecs = -ts->nsecs;
negative = TRUE;
} else {
/*
* -2147483648 is the smallest number that fits in a signed
* 2's complement 32-bit variable, and its negative doesn't
* fit in 32 bits.
*
* Just cast it to a 32-bit unsigned int value to set the
* 32-bit unsigned int variable to 2147483648.
*
* Note that, on platforms where both integers and long
* integers are 32-bit, such as 32-bit UN*Xes and both
* 32-bit *and* 64-bit Windows, making the variable in
* question a long will not avoid undefined behavior.
*/
nsecs = (guint32)ts->nsecs;
negative = TRUE;
}
ptr = buf;
remaining = COL_MAX_LEN;
if (secs >= (60*60)) {
num_bytes = snprintf(ptr, remaining, "%s%dh %2dm %2d",
negative ? "- " : "",
(gint32) secs / (60 * 60),
(gint32) (secs / 60) % 60,
(gint32) secs % 60);
} else if (secs >= 60) {
num_bytes = snprintf(ptr, remaining, "%s%dm %2d",
negative ? "- " : "",
(gint32) secs / 60,
(gint32) secs % 60);
} else {
num_bytes = snprintf(ptr, remaining, "%s%d",
negative ? "- " : "",
(gint32) secs);
}
if (num_bytes < 0) {
/*
* That got an error.
* Not much else we can do.
*/
snprintf(ptr, remaining, "snprintf() failed");
return;
}
if ((unsigned int)num_bytes >= remaining) {
/*
* That filled up or would have overflowed the buffer.
* Nothing more we can do.
*/
return;
}
ptr += num_bytes;
remaining -= num_bytes;
tsprecision = get_frame_timestamp_precision(fd);
if (tsprecision != 0) {
/*
* Append the fractional part.
*/
num_bytes = format_fractional_part_nsecs(ptr, remaining, nsecs, decimal_point, tsprecision);
if ((unsigned int)num_bytes >= remaining) {
/*
* That filled up or would have overflowed the buffer.
* Nothing more we can do.
*/
return;
}
ptr += num_bytes;
remaining -= num_bytes;
}
/* Append the "s" for seconds. */
snprintf(ptr, remaining, "s");
}
static void
col_set_rel_time(const frame_data *fd, column_info *cinfo, const int col)
{
nstime_t del_rel_ts;
if (!fd->has_ts) {
cinfo->columns[col].col_buf[0] = '\0';
return;
}
frame_delta_abs_time(cinfo->epan, fd, fd->frame_ref_num, &del_rel_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_rel_ts, cinfo->columns[col].col_buf);
cinfo->col_expr.col_expr[col] = "frame.time_relative";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_hour_min_sec(fd, &del_rel_ts, cinfo->columns[col].col_buf, col_decimal_point);
cinfo->col_expr.col_expr[col] = "frame.time_relative";
set_time_seconds(fd, &del_rel_ts, cinfo->col_expr.col_expr_val[col]);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
col_set_delta_time(const frame_data *fd, column_info *cinfo, const int col)
{
nstime_t del_cap_ts;
if (!fd->has_ts) {
cinfo->columns[col].col_buf[0] = '\0';
return;
}
frame_delta_abs_time(cinfo->epan, fd, fd->num - 1, &del_cap_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_cap_ts, cinfo->columns[col].col_buf);
cinfo->col_expr.col_expr[col] = "frame.time_delta";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_hour_min_sec(fd, &del_cap_ts, cinfo->columns[col].col_buf, col_decimal_point);
cinfo->col_expr.col_expr[col] = "frame.time_delta";
set_time_seconds(fd, &del_cap_ts, cinfo->col_expr.col_expr_val[col]);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
col_set_delta_time_dis(const frame_data *fd, column_info *cinfo, const int col)
{
nstime_t del_dis_ts;
if (!fd->has_ts) {
cinfo->columns[col].col_buf[0] = '\0';
return;
}
frame_delta_abs_time(cinfo->epan, fd, fd->prev_dis_num, &del_dis_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_dis_ts, cinfo->columns[col].col_buf);
cinfo->col_expr.col_expr[col] = "frame.time_delta_displayed";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_hour_min_sec(fd, &del_dis_ts, cinfo->columns[col].col_buf, col_decimal_point);
cinfo->col_expr.col_expr[col] = "frame.time_delta_displayed";
set_time_seconds(fd, &del_dis_ts, cinfo->col_expr.col_expr_val[col]);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
/*
* Time, without date.
*/
static void
set_abs_time(const frame_data *fd, gchar *buf, char *decimal_point, gboolean local)
{
struct tm tm, *tmp;
gchar *ptr;
size_t remaining;
int num_bytes;
int tsprecision;
if (!fd->has_ts) {
*buf = '\0';
return;
}
ptr = buf;
remaining = COL_MAX_LEN;
if (local)
tmp = ws_localtime_r(&fd->abs_ts.secs, &tm);
else
tmp = ws_gmtime_r(&fd->abs_ts.secs, &tm);
if (tmp == NULL) {
snprintf(ptr, remaining, "Not representable");
return;
}
/* Integral part. */
num_bytes = snprintf(ptr, remaining, "%02d:%02d:%02d",
tmp->tm_hour,
tmp->tm_min,
tmp->tm_sec);
if (num_bytes < 0) {
/*
* That got an error.
* Not much else we can do.
*/
snprintf(ptr, remaining, "snprintf() failed");
return;
}
if ((unsigned int)num_bytes >= remaining) {
/*
* That filled up or would have overflowed the buffer.
* Nothing more we can do.
*/
return;
}
ptr += num_bytes;
remaining -= num_bytes;
tsprecision = get_frame_timestamp_precision(fd);
if (tsprecision != 0) {
/*
* Append the fractional part.
* Get the nsecs as a 32-bit unsigned value, as it should never
* be negative, so we treat it as unsigned.
*/
format_fractional_part_nsecs(ptr, remaining, (guint32)fd->abs_ts.nsecs, decimal_point, tsprecision);
}
}
static void
col_set_abs_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_time(fd, cinfo->columns[col].col_buf, col_decimal_point, TRUE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static void
col_set_utc_time(const frame_data *fd, column_info *cinfo, const int col)
{
set_abs_time(fd, cinfo->columns[col].col_buf, col_decimal_point, FALSE);
cinfo->col_expr.col_expr[col] = "frame.time";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
static gboolean
set_epoch_time(const frame_data *fd, gchar *buf)
{
if (!fd->has_ts) {
buf[0] = '\0';
return FALSE;
}
display_epoch_time(buf, COL_MAX_LEN, &fd->abs_ts, get_frame_timestamp_precision(fd));
return TRUE;
}
static void
col_set_epoch_time(const frame_data *fd, column_info *cinfo, const int col)
{
if (set_epoch_time(fd, cinfo->columns[col].col_buf)) {
cinfo->col_expr.col_expr[col] = "frame.time_delta";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],cinfo->columns[col].col_buf,COL_MAX_LEN);
}
cinfo->columns[col].col_data = cinfo->columns[col].col_buf;
}
void
set_fd_time(const epan_t *epan, frame_data *fd, gchar *buf)
{
switch (timestamp_get_type()) {
case TS_ABSOLUTE:
set_abs_time(fd, buf, col_decimal_point, TRUE);
break;
case TS_ABSOLUTE_WITH_YMD:
set_abs_ymd_time(fd, buf, col_decimal_point, TRUE);
break;
case TS_ABSOLUTE_WITH_YDOY:
set_abs_ydoy_time(fd, buf, col_decimal_point, TRUE);
break;
case TS_RELATIVE:
if (fd->has_ts) {
nstime_t del_rel_ts;
frame_delta_abs_time(epan, fd, fd->frame_ref_num, &del_rel_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_rel_ts, buf);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_seconds(fd, &del_rel_ts, buf);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
} else {
buf[0] = '\0';
}
break;
case TS_DELTA:
if (fd->has_ts) {
nstime_t del_cap_ts;
frame_delta_abs_time(epan, fd, fd->num - 1, &del_cap_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_cap_ts, buf);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_hour_min_sec(fd, &del_cap_ts, buf, col_decimal_point);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
} else {
buf[0] = '\0';
}
break;
case TS_DELTA_DIS:
if (fd->has_ts) {
nstime_t del_dis_ts;
frame_delta_abs_time(epan, fd, fd->prev_dis_num, &del_dis_ts);
switch (timestamp_get_seconds_type()) {
case TS_SECONDS_DEFAULT:
set_time_seconds(fd, &del_dis_ts, buf);
break;
case TS_SECONDS_HOUR_MIN_SEC:
set_time_hour_min_sec(fd, &del_dis_ts, buf, col_decimal_point);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
} else {
buf[0] = '\0';
}
break;
case TS_EPOCH:
set_epoch_time(fd, buf);
break;
case TS_UTC:
set_abs_time(fd, buf, col_decimal_point, FALSE);
break;
case TS_UTC_WITH_YMD:
set_abs_ymd_time(fd, buf, col_decimal_point, FALSE);
break;
case TS_UTC_WITH_YDOY:
set_abs_ydoy_time(fd, buf, col_decimal_point, FALSE);
break;
case TS_NOT_SET:
/* code is missing for this case, but I don't know which [jmayer20051219] */
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
break;
}
}
static void
col_set_cls_time(const frame_data *fd, column_info *cinfo, const gint col)
{
switch (timestamp_get_type()) {
case TS_ABSOLUTE:
col_set_abs_time(fd, cinfo, col);
break;
case TS_ABSOLUTE_WITH_YMD:
col_set_abs_ymd_time(fd, cinfo, col);
break;
case TS_ABSOLUTE_WITH_YDOY:
col_set_abs_ydoy_time(fd, cinfo, col);
break;
case TS_RELATIVE:
col_set_rel_time(fd, cinfo, col);
break;
case TS_DELTA:
col_set_delta_time(fd, cinfo, col);
break;
case TS_DELTA_DIS:
col_set_delta_time_dis(fd, cinfo, col);
break;
case TS_EPOCH:
col_set_epoch_time(fd, cinfo, col);
break;
case TS_UTC:
col_set_utc_time(fd, cinfo, col);
break;
case TS_UTC_WITH_YMD:
col_set_utc_ymd_time(fd, cinfo, col);
break;
case TS_UTC_WITH_YDOY:
col_set_utc_ydoy_time(fd, cinfo, col);
break;
case TS_NOT_SET:
/* code is missing for this case, but I don't know which [jmayer20051219] */
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
break;
}
}
/* Set the format of the variable time format. */
static void
col_set_fmt_time(const frame_data *fd, column_info *cinfo, const gint fmt, const gint col)
{
COL_CHECK_REF_TIME(fd, cinfo->columns[col].col_buf);
switch (fmt) {
case COL_CLS_TIME:
col_set_cls_time(fd, cinfo, col);
break;
case COL_ABS_TIME:
col_set_abs_time(fd, cinfo, col);
break;
case COL_ABS_YMD_TIME:
col_set_abs_ymd_time(fd, cinfo, col);
break;
case COL_ABS_YDOY_TIME:
col_set_abs_ydoy_time(fd, cinfo, col);
break;
case COL_REL_TIME:
col_set_rel_time(fd, cinfo, col);
break;
case COL_DELTA_TIME:
col_set_delta_time(fd, cinfo, col);
break;
case COL_DELTA_TIME_DIS:
col_set_delta_time_dis(fd, cinfo, col);
break;
case COL_UTC_TIME:
col_set_utc_time(fd, cinfo, col);
break;
case COL_UTC_YMD_TIME:
col_set_utc_ymd_time(fd, cinfo, col);
break;
case COL_UTC_YDOY_TIME:
col_set_utc_ydoy_time(fd, cinfo, col);
break;
default:
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
break;
}
}
/* --------------------------- */
/* Set the given (relative) time to a column element.
*
* Used by dissectors to set the time in a column
*
* @param cinfo the current packet row
* @param el the column to use, e.g. COL_INFO
* @param ts the time to set in the column
* @param fieldname the fieldname to use for creating a filter (when
* applying/preparing/copying as filter)
*/
void
col_set_time(column_info *cinfo, const gint el, const nstime_t *ts, const char *fieldname)
{
int col;
col_item_t* col_item;
if (!CHECK_COL(cinfo, el))
return;
/** @todo TODO: We don't respect fd->ref_time (no way to access 'fd')
COL_CHECK_REF_TIME(fd, buf);
*/
for (col = cinfo->col_first[el]; col <= cinfo->col_last[el]; col++) {
col_item = &cinfo->columns[col];
if (col_item->fmt_matx[el]) {
display_signed_time(col_item->col_buf, COL_MAX_LEN, ts, get_default_timestamp_precision());
col_item->col_data = col_item->col_buf;
cinfo->col_expr.col_expr[col] = fieldname;
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col],col_item->col_buf,COL_MAX_LEN);
}
}
}
static void
col_set_addr(packet_info *pinfo, const int col, const address *addr, const gboolean is_src,
const gboolean fill_col_exprs, const gboolean res)
{
const char *name;
col_item_t* col_item = &pinfo->cinfo->columns[col];
if (addr->type == AT_NONE) {
/* No address, nothing to do */
return;
}
if (res && (name = address_to_name(addr)) != NULL)
col_item->col_data = name;
else {
col_item->col_data = col_item->col_buf;
address_to_str_buf(addr, col_item->col_buf, COL_MAX_LEN);
}
if (!fill_col_exprs)
return;
pinfo->cinfo->col_expr.col_expr[col] = address_type_column_filter_string(addr, is_src);
/* For address types that have a filter, create a string */
if (strlen(pinfo->cinfo->col_expr.col_expr[col]) > 0)
address_to_str_buf(addr, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
}
/* ------------------------ */
static void
col_set_port(packet_info *pinfo, const int col, const gboolean is_res, const gboolean is_src, const gboolean fill_col_exprs _U_)
{
guint32 port;
col_item_t* col_item = &pinfo->cinfo->columns[col];
if (is_src)
port = pinfo->srcport;
else
port = pinfo->destport;
/* TODO: Use fill_col_exprs */
switch (pinfo->ptype) {
case PT_SCTP:
if (is_res)
(void) g_strlcpy(col_item->col_buf, sctp_port_to_display(pinfo->pool, port), COL_MAX_LEN);
else
guint32_to_str_buf(port, col_item->col_buf, COL_MAX_LEN);
break;
case PT_TCP:
guint32_to_str_buf(port, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
if (is_res)
(void) g_strlcpy(col_item->col_buf, tcp_port_to_display(pinfo->pool, port), COL_MAX_LEN);
else
(void) g_strlcpy(col_item->col_buf, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "tcp.srcport";
else
pinfo->cinfo->col_expr.col_expr[col] = "tcp.dstport";
break;
case PT_UDP:
guint32_to_str_buf(port, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
if (is_res)
(void) g_strlcpy(col_item->col_buf, udp_port_to_display(pinfo->pool, port), COL_MAX_LEN);
else
(void) g_strlcpy(col_item->col_buf, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "udp.srcport";
else
pinfo->cinfo->col_expr.col_expr[col] = "udp.dstport";
break;
case PT_DDP:
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "ddp.src_socket";
else
pinfo->cinfo->col_expr.col_expr[col] = "ddp.dst_socket";
guint32_to_str_buf(port, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
(void) g_strlcpy(col_item->col_buf, pinfo->cinfo->col_expr.col_expr_val[col], COL_MAX_LEN);
break;
case PT_IPX:
/* XXX - resolve IPX socket numbers */
snprintf(col_item->col_buf, COL_MAX_LEN, "0x%04x", port);
(void) g_strlcpy(pinfo->cinfo->col_expr.col_expr_val[col], col_item->col_buf,COL_MAX_LEN);
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "ipx.src.socket";
else
pinfo->cinfo->col_expr.col_expr[col] = "ipx.dst.socket";
break;
case PT_IDP:
/* XXX - resolve IDP socket numbers */
snprintf(col_item->col_buf, COL_MAX_LEN, "0x%04x", port);
(void) g_strlcpy(pinfo->cinfo->col_expr.col_expr_val[col], col_item->col_buf,COL_MAX_LEN);
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "idp.src.socket";
else
pinfo->cinfo->col_expr.col_expr[col] = "idp.dst.socket";
break;
case PT_USB:
/* XXX - resolve USB endpoint numbers */
snprintf(col_item->col_buf, COL_MAX_LEN, "0x%08x", port);
(void) g_strlcpy(pinfo->cinfo->col_expr.col_expr_val[col], col_item->col_buf,COL_MAX_LEN);
if (is_src)
pinfo->cinfo->col_expr.col_expr[col] = "usb.src.endpoint";
else
pinfo->cinfo->col_expr.col_expr[col] = "usb.dst.endpoint";
break;
default:
break;
}
col_item->col_data = col_item->col_buf;
}
gboolean
col_based_on_frame_data(column_info *cinfo, const gint col)
{
2021-06-18 18:21:42 +00:00
ws_assert(cinfo);
ws_assert(col < cinfo->num_cols);
switch (cinfo->columns[col].col_fmt) {
case COL_NUMBER:
case COL_CLS_TIME:
case COL_ABS_TIME:
case COL_ABS_YMD_TIME:
case COL_ABS_YDOY_TIME:
case COL_UTC_TIME:
case COL_UTC_YMD_TIME:
case COL_UTC_YDOY_TIME:
case COL_REL_TIME:
case COL_DELTA_TIME:
case COL_DELTA_TIME_DIS:
case COL_PACKET_LENGTH:
case COL_CUMULATIVE_BYTES:
return TRUE;
default:
return FALSE;
}
}
void
col_fill_in_frame_data(const frame_data *fd, column_info *cinfo, const gint col, const gboolean fill_col_exprs)
{
col_item_t* col_item = &cinfo->columns[col];
switch (col_item->col_fmt) {
case COL_NUMBER:
guint32_to_str_buf(fd->num, col_item->col_buf, COL_MAX_LEN);
col_item->col_data = col_item->col_buf;
break;
case COL_CLS_TIME:
case COL_ABS_TIME:
case COL_ABS_YMD_TIME:
case COL_ABS_YDOY_TIME:
case COL_UTC_TIME:
case COL_UTC_YMD_TIME:
case COL_UTC_YDOY_TIME:
case COL_REL_TIME:
case COL_DELTA_TIME:
case COL_DELTA_TIME_DIS:
/* TODO: Pass on fill_col_exprs */
col_set_fmt_time(fd, cinfo, col_item->col_fmt, col);
break;
case COL_PACKET_LENGTH:
guint32_to_str_buf(fd->pkt_len, col_item->col_buf, COL_MAX_LEN);
col_item->col_data = col_item->col_buf;
break;
case COL_CUMULATIVE_BYTES:
guint32_to_str_buf(fd->cum_bytes, col_item->col_buf, COL_MAX_LEN);
col_item->col_data = col_item->col_buf;
break;
default:
break;
}
if (!fill_col_exprs)
return;
switch (col_item->col_fmt) {
case COL_NUMBER:
cinfo->col_expr.col_expr[col] = "frame.number";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col], col_item->col_buf, COL_MAX_LEN);
break;
case COL_CLS_TIME:
case COL_ABS_TIME:
case COL_ABS_YMD_TIME:
case COL_ABS_YDOY_TIME:
case COL_UTC_TIME:
case COL_UTC_YMD_TIME:
case COL_UTC_YDOY_TIME:
case COL_REL_TIME:
case COL_DELTA_TIME:
case COL_DELTA_TIME_DIS:
/* Already handled above */
break;
case COL_PACKET_LENGTH:
cinfo->col_expr.col_expr[col] = "frame.len";
(void) g_strlcpy(cinfo->col_expr.col_expr_val[col], col_item->col_buf, COL_MAX_LEN);
break;
case COL_CUMULATIVE_BYTES:
break;
default:
break;
}
}
void
col_fill_in(packet_info *pinfo, const gboolean fill_col_exprs, const gboolean fill_fd_colums)
{
int i;
col_item_t* col_item;
if (!pinfo->cinfo)
return;
for (i = 0; i < pinfo->cinfo->num_cols; i++) {
col_item = &pinfo->cinfo->columns[i];
if (col_based_on_frame_data(pinfo->cinfo, i)) {
if (fill_fd_colums)
col_fill_in_frame_data(pinfo->fd, pinfo->cinfo, i, fill_col_exprs);
} else {
switch (col_item->col_fmt) {
case COL_DEF_SRC:
case COL_RES_SRC: /* COL_DEF_SRC is currently just like COL_RES_SRC */
col_set_addr(pinfo, i, &pinfo->src, TRUE, fill_col_exprs, TRUE);
break;
case COL_UNRES_SRC:
col_set_addr(pinfo, i, &pinfo->src, TRUE, fill_col_exprs, FALSE);
break;
case COL_DEF_DL_SRC:
case COL_RES_DL_SRC:
col_set_addr(pinfo, i, &pinfo->dl_src, TRUE, fill_col_exprs, TRUE);
break;
case COL_UNRES_DL_SRC:
col_set_addr(pinfo, i, &pinfo->dl_src, TRUE, fill_col_exprs, FALSE);
break;
case COL_DEF_NET_SRC:
case COL_RES_NET_SRC:
col_set_addr(pinfo, i, &pinfo->net_src, TRUE, fill_col_exprs, TRUE);
break;
case COL_UNRES_NET_SRC:
col_set_addr(pinfo, i, &pinfo->net_src, TRUE, fill_col_exprs, FALSE);
break;
case COL_DEF_DST:
case COL_RES_DST: /* COL_DEF_DST is currently just like COL_RES_DST */
col_set_addr(pinfo, i, &pinfo->dst, FALSE, fill_col_exprs, TRUE);
break;
case COL_UNRES_DST:
col_set_addr(pinfo, i, &pinfo->dst, FALSE, fill_col_exprs, FALSE);
break;
case COL_DEF_DL_DST:
case COL_RES_DL_DST:
col_set_addr(pinfo, i, &pinfo->dl_dst, FALSE, fill_col_exprs, TRUE);
break;
case COL_UNRES_DL_DST:
col_set_addr(pinfo, i, &pinfo->dl_dst, FALSE, fill_col_exprs, FALSE);
break;
case COL_DEF_NET_DST:
case COL_RES_NET_DST:
col_set_addr(pinfo, i, &pinfo->net_dst, FALSE, fill_col_exprs, TRUE);
break;
case COL_UNRES_NET_DST:
col_set_addr(pinfo, i, &pinfo->net_dst, FALSE, fill_col_exprs, FALSE);
break;
case COL_DEF_SRC_PORT:
case COL_RES_SRC_PORT: /* COL_DEF_SRC_PORT is currently just like COL_RES_SRC_PORT */
col_set_port(pinfo, i, TRUE, TRUE, fill_col_exprs);
break;
case COL_UNRES_SRC_PORT:
col_set_port(pinfo, i, FALSE, TRUE, fill_col_exprs);
break;
case COL_DEF_DST_PORT:
case COL_RES_DST_PORT: /* COL_DEF_DST_PORT is currently just like COL_RES_DST_PORT */
col_set_port(pinfo, i, TRUE, FALSE, fill_col_exprs);
break;
case COL_UNRES_DST_PORT:
col_set_port(pinfo, i, FALSE, FALSE, fill_col_exprs);
break;
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
case COL_CUSTOM:
/* Formatting handled by col_custom_set_edt() / col_custom_get_filter() */
break;
case NUM_COL_FMTS: /* keep compiler happy - shouldn't get here */
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
break;
default:
if (col_item->col_fmt >= NUM_COL_FMTS) {
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
/*
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
* Formatting handled by expert.c (COL_EXPERT), or individual
* dissectors. Fill in from the text using the internal hfid.
*/
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
if (fill_col_exprs) {
pinfo->cinfo->col_expr.col_expr[i] = proto_registrar_get_nth(col_item->hf_id)->abbrev;
(void) g_strlcpy(pinfo->cinfo->col_expr.col_expr_val[i], pinfo->cinfo->columns[i].col_data, (col_item->col_fmt == COL_INFO) ? COL_MAX_INFO_LEN : COL_MAX_LEN);
}
break;
}
}
}
}
/*
* Fill in columns if we got an error reading the packet.
* We set most columns to "???", fill in columns that don't need data read
* from the file, and set the Info column to an error message.
*/
void
col_fill_in_error(column_info *cinfo, frame_data *fdata, const gboolean fill_col_exprs, const gboolean fill_fd_colums)
{
int i;
col_item_t* col_item;
if (!cinfo)
return;
for (i = 0; i < cinfo->num_cols; i++) {
col_item = &cinfo->columns[i];
if (col_based_on_frame_data(cinfo, i)) {
if (fill_fd_colums)
col_fill_in_frame_data(fdata, cinfo, i, fill_col_exprs);
} else if (col_item->col_fmt == COL_INFO) {
/* XXX - say more than this */
col_item->col_data = "Read error";
} else {
if (col_item->col_fmt >= NUM_COL_FMTS) {
2021-06-18 18:21:42 +00:00
ws_assert_not_reached();
}
/*
* No dissection was done, and these columns are set as the
* result of the dissection, so....
*/
col_item->col_data = "???";
break;
}
}
}
gboolean col_data_changed(void) {
gboolean cur_cdc = col_data_changed_;
col_data_changed_ = FALSE;
return cur_cdc;
}
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
void
col_register_protocol(void)
{
/* This gets called by proto_init() before column_register_fields()
* gets called by the preference modules actually getting registered.
*/
if (proto_cols <= 0) {
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
proto_cols = proto_get_id_by_filter_name("_ws.col");
}
if (proto_cols <= 0) {
epan: Register dynamic column fields and make them filterable Make the text of each registered column a FT_STRING field that can be filtered, prefixed with _ws.col - these work in display filters, filters in taps, coloring rules, Wireshark read filters, and in the -Y, -R, -e, and -j options to tshark. Use them as the default "Apply as Filter" value for the columns that aren't handled by anything else currently. Because only the columns formats that actually correspond to columns get filled in (invisible columns work), register and deregister the fields when the columns change. Use the lower case version of the rest of the COL_* define for each column as the field name. This adds a number of conditions to "when are the columns needed", including when the main display filter or any filter on a tap is using one of these fields. Custom columns are currently not implemented. For custom columns, the tree then has to be further primed with any fields used by the custom columns as well. (Perhaps that should happen in epan_dissect_run() - are there any cases where we construct the columns and don't want to prime with any field that custom columns contains? Possibly in taps that we know only use build in columns.) Thus, for performance reasons, you're better off matching an ordinary field if possible; it takes extra time to generate the columns and many of them are numeric types. (Note that you can always convert a non-string field to a string field if you want regex matching, consult the *wireshark-filter(4)* man page.) It does save a bit on typing (especially for a multifield custom column) and remembering the column title might be easier in some cases. The columns are set before the color filters, which means that you can have a color filter that depends on a built-in column like Info or Protocol. Remove the special handling for the -e option to tshark. Note that the behavior is a little different now, because fixed field names are used instead of the titles (using the titles allowed illegal filter names, because it wasn't going through the filter engine.) For default names, this means that they're no longer capitalized, so "_ws.col.info" instead of "_ws.col.Info" - hopefully a small price in exchange for the filters working everywhere. The output format for -T fields remains the same; all that special handling is removed (except for remembering if someone asked for a column field to know that columns should be constructed.) They're also set before the postdissectors, so postdissectors can have access. Anything that depends on whether a packet and previous packets are displayed (COL_DELTA_TIME_DIS or COL_CUMULATIVE_BYTES) doesn't work the way most people expect, so don't register fields for those. (The same is already true of color filters that use those, along with color filters that use the color filter fields.) Fix #16576. Fix #17971. Fix #4684. Fix #13491. Fix #13941.
2023-05-01 12:02:18 +00:00
proto_cols = proto_register_protocol("Wireshark Columns", "Columns", "_ws.col");
}
static gint *ett[] = {
&ett_cols
};
proto_register_subtree_array(ett, G_N_ELEMENTS(ett));
}
void
col_dissect(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree)
{
proto_item *ti;
proto_tree *col_tree;
column_info *cinfo = pinfo->cinfo;
if (!cinfo) {
return;
}
if (proto_field_is_referenced(tree, proto_cols)) {
// XXX: Needed if we also create _ws.col.custom
//col_custom_set(tree, cinfo);
col_fill_in(pinfo, FALSE, TRUE);
ti = proto_tree_add_item(tree, proto_cols, tvb, 0, 0, ENC_NA);
proto_item_set_hidden(ti);
col_tree = proto_item_add_subtree(ti, ett_cols);
for (int i = 0; i < cinfo->num_cols; ++i) {
if (cinfo->columns[i].hf_id != -1) {
if (cinfo->columns[i].col_fmt == COL_CUSTOM) {
ti = proto_tree_add_string_format(col_tree, cinfo->columns[i].hf_id, tvb, 0, 0, get_column_text(cinfo, i), "%s: %s", get_column_title(i), get_column_text(cinfo, i));
} else {
ti = proto_tree_add_string(col_tree, cinfo->columns[i].hf_id, tvb, 0, 0, get_column_text(cinfo, i));
}
proto_item_set_hidden(ti);
}
}
}
}
/*
* Editor modelines
*
* Local Variables:
* c-basic-offset: 2
* tab-width: 8
* indent-tabs-mode: nil
* End:
*
* ex: set shiftwidth=2 tabstop=8 expandtab:
* :indentSize=2:tabSize=8:noTabs=true:
*/