Prior to this patch, any SwMatchError raised within the 'transport'
would not be interpreted.
EXCEPTION of type 'SwMatchError' occurred with message: 'SW match failed! Expected 9000 and got 6982.'
vs (now)
EXCEPTION of type 'SwMatchError' occurred with message: 'SW match failed! Expected 9000 and got 6982: Command not allowed - Security status not satisfied'
Change-Id: I08b7f2b6bd422f7f2f36094bc8a29b187ff882a6
* move existing docs to sphinx / autodoc
* add more api documentation
* improve wording on some exception strings
Change-Id: Ia41e14d643d452d92fc8d3c2fb9c4ac9021402e9
"data" is an awfully generic term. Anything stored on a card is data.
This specific code deals with resolving key/pin material from an
external source.
Change-Id: I4c8e1be3e766f7c0565c07b39d48abf8adc375af
As we can notice during 'export': Some files had been defined
as LinFixed but are Transparent - and vice versa. Let's fix those
an bring our definitions in sync with the specs.
Change-Id: I365ece7b82a1c79b3af87a79ff964d7989362789
When the CardFile hierarchy talks about 'application' it means CardADF.
When the RuntimeState and CardProfile talk about 'application' they mean
a CardApplication.
Let's clarify this in the file names, and make CardADF have an optional
reference to the CardApplication, so that application specific status
word interpretation really works.
Change-Id: Ibc80a41d79dca547f14d5d84f447742e6b46d7ca
* add type annotations in-line with PEP484
* convert existing documentation to follow the
"Google Python Style Guide" format understood by
the sphinx.ext.napoleon' extension
* add much more documentation all over the code base
Change-Id: I6ac88e0662cf3c56ae32d86d50b18a8b4150571a
In Change-Id I848a766e6d00be497c7db905475e0681cce197ac we added a CardDF
instance for DF_5GS. That DF should not have provided a
decode_select_response() method, and instead fall back to that of the
base class, which calls the method of the parent directory (ADF_USIM).
The difference is illustrated below
pySIM-shell (MF/ADF.USIM/EF.IMSI)> select DF.5GS
"622e8202782183025fc0a509800171830400018d088a01058c056611111111c60f90017083010183018183010a83010b"
vs. (with this patch):
pySIM-shell (MF/ADF.USIM)> select DF.5GS
{
"file_descriptor": {
"shareable": true,
"file_type": "df",
"structure": "no_info_given"
},
"file_identifier": "5FC0",
"proprietary_info": {
"uicc_characteristics": "71",
"available_memory": 101640
},
"life_cycle_status_int": "operational_activated",
"security_attrib_compact": "6611111111",
"pin_status_template_do": "90017083010183018183010A83010B"
}
Change-Id: I80612711bbc8c47285a828a0759b20beea6619f1
At the moment we only have a basic version of a verify_chv commnad, but
in order to handle any CHV/PIN related situation we also need commands
to enable, disable, change and unblock CHV.
- fix verify_chv commnad: more distinct parameter names, better help
strings, correct pin code encoding and add external source lookup
- Add unblock_chv, change_chv, enable_chv and disable_chv commands
- add/fix related functions in commands.py
Change-Id: Ic89446e6bd2021095e579fb6b20458df48ba6413
Related: OS#4963
When a record or a binary file is written the card goes throth a full
flash/eeprom write cycle at this location, even when the data does not
change. This can be optimized by reading before writing in order to
compere if the data we are about to write is actually different.
Change-Id: Ifd1b80d3ede15a7caa29077a37ac7cf58c9053f1
Related: OS#4963
It can be hard to manage ADM pins when working with different cards at
the same time. To make this easier, add an automatic way to determine
the ADM pin for each card from a CSV file.
- add a CardData clas model that can be extended to to get the data from
various different sources. For now use CSV-Files. Also add a way how
multiple CardData classes can be registered so that one global get
function can query all registered CardData classes at once.
- automatically check for CSV-File in home directory and use it as
default CardData source unless the user specifies a CSV file via
commandline argument.
- extend the verify_adm command so that it automatically queries the
ADM pin if no argument is given. Also do not try to authenticate if
no ADM pin could be determined.
Change-Id: I51835ccb16bcbce35e7f3765e8927a4451509e77
Related: OS#4963
When the ADF is selected, then this is done by the AID. At the moment
only the first 7 bytes of the AID are used to select the ADF.
sysmo-isim-sja2 tolerates this, but sysmo-usim-sjs1 does not. The Cards
class already has methods to deal with this problem. The method
select_adf_by_aid takes an ADF name and completes the AID from an
internal list. This can be extended to support partial hexadecimal AIDs
as well.
Change-Id: If99b143ae5ff42a889c52e8023084692e709e1b1
Related: OS#4963
The class ShellCommands defined in ADF_USIM overloads useful CommandSet
classes defined in the superclass, making their commands inaccessible.
Also ts_31_102 does not have such a class definition in the ADF_ISIM
class, so lets remove this class.
Change-Id: I0e67c570fc4f17641d990a9cd239632ecf622de3
Related: OS#4963
Some cards may have additional propritary EF files which pySim-shell
does not support. If the user knows the exact FID the file can still be
selected and it is possible to read the file type and memory model from
the select response. This info can be used to create a new file object
at runtime that will work like any other EF/DF.
Change-Id: Iafff97443130f8bb8c5bc68f51d2fe1d93fff07c
Related: OS#4963
The Change I83d718ff9c3ff6aef47930f38d7f50424f9b880f removes the
keyword arguments from the CardProfile class constructor. This requires
us to use the keywords during instantiation since we can not rely on
the position anymore.
Change-Id: Ia62597c59287848662dbbedcc38ba90f183c4aca
The do_update_... functions do always print the returned data. However,
there may be no data. If this is the case than an empty line is printed.
This may cause ugly log output, especially when working with scripts.
Change-Id: Ia9715d46ec957544cfbeea98d2fe15eb74f5b884
Related: OS#4963
Having lists and dictionaries as default argument values is a bad
idea, because the same instance of list/dict will be used by all
objects instantiated using such constructor:
def appendItem(itemName, itemList=[]):
itemList.append(itemName)
return itemList
print(appendItem('notebook'))
print(appendItem('pencil'))
print(appendItem('eraser'))
Output:
['notebook']
['notebook', 'pencil']
['notebook', 'pencil', 'eraser']
Change-Id: I83d718ff9c3ff6aef47930f38d7f50424f9b880f
currently ADF.ISIM and ADF.USIM are always added regardless if there is
a matching application on the card or not. Lets check what applications
are actually installed and add ADF.ISIM and ADF.USIM dynamically.
Change-Id: I42ee23375f98e6322708c1c4db6d65e1425feecd
Related: OS#4963
In the method add_application() the method name should be append()
instead of add().
Change-Id: Ic8ad62567968e09786eac86f219b56a3d3200511
Related: OS#4963
The SW_match function takes a given status word and compares it against
a pattern that may contain wildcards (x or ?). This works by creating a
masked version of the SW using a pattern first (each hex digit is
replaced by a wildcard charafter if the pattern has a wildcard in the
same position). Once this is done, the resulting masked version is
compared at the pattern. However, the current implementation can not
work, since it compares the input SW against the pattern to decide
wihich chrafters should be masked. The input SW never contains wildcard
charafters.
Change-Id: I805ad32160fcfcb8628bf919b64f7eee0fe03c7e
Related: OS#4963
The _scc.veryif_adm() method already does status word checking
internally and also raises an execption should the authentication be
unsuccessful, so we do not have to put an additional status word check +
execition when we use the method from cards.
Change-Id: I785d27e4d49a9cda1a771b56ce5ac9c1f1d1e79a
Related: OS#4963
At the moment we use the send_apdu_checksw() method to send the APDU for
ADM authentication. This method only checks if the command returns with
sw = 9000. If not it raises an exception that the sw is not as expected.
The user may think that this is a problem with thr reader, pcscd or
pySim in the first place and may try multiple times until the card is
permanently locked. A better execption string that also displays the
tries which are left may be helpful.
Change-Id: Icf428831094f8c1045eefaa8cb2b92e6a36b0c13
Related: OS#4963
The file identifier of a file is strictly defined as a two digit
hexadecimal number. Do not allow adding child files that violate this
constraint.
Change-Id: I096907285b742e611d221b03ba067ea2522e7e52
Related: OS#4963
The __main__ function in filesystem.py seems to be some experimental
testcode from the very beginning of pySim-shell. Lets drop it.
Change-Id: I34f459469dfc45711ad0928c83184d7f99e0f5e3
Related: OS#4963
The method add_file of class CardDF does some constraint checking
to the basic file parameters (e.g. fid). Since one might also expect
those checks in the superclass CardFile lets leave a comment to make
the code better understandable.
Change-Id: Iebae28909fe6aade3bd4024112a222819572d735
Related: OS#4963
It is better to use the term "fid" instead of "name" when a reserved FID
is detected.
Change-Id: I054f3b3a156f0164c62610cfde1aec2145c20925
Related: OS#4963
pysim-shell does not have a convinient way to list the files available
in one directory. Tab completion currently is the only way to obtain a
list of the available files. Lets add a dir command to print a file
list.
Change-Id: Ic06a60e0a0ec54d9bb26e151453ceb49d69e3df7
Related: OS#4963
The flags NAMES, FIDS and APPS do not properly distinguish between
applications and normal files. With APPS it is only possible to exclude
or include the selectable applications in a list with NAMES or FIDS, but
it is not possible to get only the application names or identifiers.
- remove the APPS flag
- rename NAMES to FNAMES and make it only normal file related
- add ANAMES and relate it only to application (ADF) names
- add AIDS and relate it only to application identifiers
Change-Id: Id07e0dcbab10cd78c1b78d37319b7b0e5e83b64d
Related: OS#4963
At the moment we can only request pySim-shell to dump a specific record
of a file. However, it may be useful to dump multiple records of a
record oriented file at once.
Change-Id: Id62db2cba4e3dfb6a7b3e6be8b892c16d11a8e3e
Related: OS#4963
The file identifier (and allso application ids for ADFs), are
hexadecimal. We should be case insensitive when accepting hex
identifiers but file names should still be full matched.
Change-Id: Ibe283a108ddc9058af77c823b7222db555e1e0f6
Related: OS#4963
since we have added pySim-shell.py that has a lot of locations where the
user can enter hexadecimal data there is an increased need for input
validation. Lets add a central is_hex function that verifies hex
strings.
Change-Id: Ia29a13c9215357dd2adf141f2ef222c823f8456d
Related: OS#4963
When requesting what DF/EF/ADF are selectable it is useful to have some
control of what we do not want in the resulting list.
Change-Id: Idb50a512bfdbfdf2e98f2ce0e89928cb0ff19f5e
Related: OS#4963
The bug that was attempted to be fixed in [1] actually was in the
encoding API - pySim.utils.enc_plmn(). According to 3GPP TS 31.102,
which points to TS 24.008, the three-digit (E)HPLMN shall be encoded
as shown below (ASCII-art interpretation):
0 1 2 3 4 5 6 7
+---+---+---+---+---+---+---+---+
| MCC Digit 2 | MCC Digit 1 |
+---+---+---+---+---+---+---+---+
| MNC Digit 3 | MCC Digit 3 |
+---+---+---+---+---+---+---+---+
| MNC Digit 2 | MNC Digit 1 |
+---+---+---+---+---+---+---+---+
while pySim.utils.enc_plmn() would produce the following:
0 1 2 3 4 5 6 7
+---+---+---+---+---+---+---+---+
| MCC Digit 2 | MCC Digit 1 |
+---+---+---+---+---+---+---+---+
| MNC Digit 1 | MCC Digit 3 |
+---+---+---+---+---+---+---+---+
| MNC Digit 3 | MNC Digit 2 |
+---+---+---+---+---+---+---+---+
Initially the _decoding_ API was correct, but then got changed in
[1] to follow buggy pySim's encoding API. As a result, a (E)HPLMN
programmed with pySim-prog.py would look correct if verified by
pySim-read.py, but the actual file content would be wrong.
This situation shows that our 'program-read-match' build verification
approach alone is insignificant. The lack of unit test coverage,
at least for the core parts of the project, makes it possible to have
symmetrical bugs in both encoding and decoding API parts unnoticed.
This problem was found while trying to enable dead unit tests in [3].
Change [1] that introduced a symmetrical bug is reverted in [2].
Change-Id: Ic7612502e1bb0d280133dabbcb5cb146fc6997e5
Related: [1] I799469206f87e930d8888367890babcb8ebe23a9
Related: [2] If6bf5383988ad442e275efc7c5a159327d104879
Related: [3] I4d4facfabc75187acd5238ff4d0f26022bd58f82
This reverts commit bdf3d3597b, which
broke pySim.utils.dec_mnc_from_plmn(). According to 3GPP TS 31.102,
which points to TS 24.008, the three-digit EHPLMN shall be encoded
as shown below (ASCII-art interpretation):
0 1 2 3 4 5 6 7
+---+---+---+---+---+---+---+---+
| MCC Digit 2 | MCC Digit 1 |
+---+---+---+---+---+---+---+---+
| MNC Digit 3 | MCC Digit 3 |
+---+---+---+---+---+---+---+---+
| MNC Digit 2 | MNC Digit 1 |
+---+---+---+---+---+---+---+---+
So the original implementation was correct, and we even had a unit
test for it. Most likely, the SIM card itself was programmed
incorrectly?
Makes 'testDecMNCfromPLMN_threeDigitMNC' pass again.
Change-Id: If6bf5383988ad442e275efc7c5a159327d104879
As it turns out, we had this set of unit tests since 2018, but
so far they were not executed during the build verification.
Let's fix this:
* run unittest in discovery mode for all files in 'tests/' (commented out);
* rename this file, so it can be automatically detected and executed;
* properly import the API to be tested.
Currently 2 out of 16 unit tests are failing, so we need to get
them passing first and then uncomment the unittest execution.
Change-Id: I4d4facfabc75187acd5238ff4d0f26022bd58f82
This commit fixes two problems (found by semgrep):
* "'foo' and 'bar' in list" is incorrect, because it's interpreted
as "'foo' and ('bar' in list)". Strings with a non-zero length
evaluate to True, thus it's True if at least 'bar' is present.
* Copy-pasted 'E-UTRAN NB-S1' checked two times.
The first condition is redundant, and the whole block can be
re-implemented using two independent 'if' statements.
Change-Id: Iceb66160cfb571db8879d3810c55d252c763d320
pySim has already been migrated to Python 3 in another change [1],
and the build verification has been migrated to Debian 10 with
Python 3.7. However, there is still some backwards compatibility
code left. Let's get rid of it.
[1] Ic78da9c03e99f59d142c93394051bbc2751f0205
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
Tweaked-by: Vadim Yanitskiy <vyanitskiy@sysmocom.de>
Change-Id: I430d173535e0cd5bb895b9dfc9070cbc40cfc8ff
Unfortunately, Debian ships old Python (3.5 vs 3.8) and old pyscard
(1.9.4 vs 1.9.9). Calling PCSCCardConnection.disconnect() from a
destructor causes warnings about ignored exceptions:
AttributeError: 'NoneType' object has no attribute 'disconnect'
AttributeError: 'NoneType' object has no attribute 'setChanged'
AttributeError: 'NoneType' object has no attribute 'SCardDisconnect'
TypeError: 'NoneType' object is not callable
All these exceptions happen in pyscard's own destructors.
Change-Id: I9c644bc5fe9791b141a30bfc13647d77937a82ee