|Name||Modified||Size||Downloads / Week||Status|
|Totals: 2 Items||20.3 kB||1|
|Author:||Ralf Schlatterbeck <firstname.lastname@example.org>|
The library includes various utility modules used in other projects.
- ast_call: A simple call manager using asterisk: This can be used to originate a call and log all event associated to that call. So you can find out the status of the call and optionally the dialstatus.
- ast_cdr: parser for CDR records in asterisk. We currently use text-files only, this might be later extended for database use.
- Autosuper magic (originally from Guido van Rossums announcement of Python 2.2): For refactoring it's a good idea that each class knows its parents (that the parent is not hardcoded in the upcall), achieved by adding a the __super attribute. Use as, e.g., self.__super.method for calling method in the parent.
- base_pickler: Allow to specify a dictionary of pickle exceptions, if the value in the dictionary is in __dict__ when pickle is called, the value in __dict__ is replaced with the value in the exception dictionary.
- bero: A configurator for the bero*fos failover switch. See my blogpost why I wrote that (short summary: to work around bugs in the firmware of the device which are not addressed by the configurator code that comes with the device).
- Config_File for python-syntax configuration files used in several of my projects.
- execute: module for executing commands and handling IO, this also has a Lock and a Log mixin. Now there is also a framework for executing processes in a pipeline, there can be fork-points in the pipeline where the output of one process feeds several pipelines. See test_exec.py and test2_exec.py.
- ETree: wrapper around ElementTree for pretty-printing and xml export. This uses delegation (not inheritance) because we can't seem to inherit from ElementTree.
- Firstname: given a string-candidate of a first name, check with popular web-site if it is really considered a first name.
- Freshmeat: Get project and release information via the freshmeat.net REST API. Also support submit of a new release. I'm now using this for announcing new releases for my projects.
- grepmime: search for pattern in email attachments (even if these are encoded)
- HTML_Parse for opening and parsing URLs. The class can be used to write a parser to extract info from a web page. After extraction the result can be pickled but we lose the ability to continue parsing after reading from the pickled result.
- inductance: Inductance calculation of air-cored cylindrical flat-winded coil according to Robert Weaver http://electronbunker.ca/CalcMethods3b.html
- IP_Address: IP v4 and v6 Addresses with subnet masking
- iter_recipes: magic with iterators
- isdn: model the status of Asterisk ISDN lines, either Linux Call Router (LCR) ISDN lines or lines managed by DAHDI. Special case for Xorcom modules which provide more information using Xorcom-specific files in /proc than native DAHDI modules.
- ldap_lib: some routines for common ldap tasks including an iterator for LDAP implementations that use paged search (notably active directory).
- Math implements Euclids algorithm for greatest common divisor and as a by-product the least common multiple
- nmap: A parser for output of nmap scans to (e.g. to generate tables in TeX)
- ocf: small framework for writing OCF scripts for the heartbeat cluster resource manager (probably also usable for the newer version called pacemaker but for now only tested with the version in Debian stable aka lenny).
- PDF_Parse for opening and parsing PDFs, needs pdftotext installed
- PDF_Signature: experimental code for checking signatures on adobe PDF documents
- Phone: Parse phone numbers (Austria-specific)
- PM_Value: Possibly missing value, encapsulate a number which may be missing and associated arithmetics.
- Rational: Implementation of Rational numbers
- stateparser: Simple state-machine parser class, this is used for writing line-based parsers using a state machine. For an example usage see nmap. After parsing the result can be pickled but we lose the ability to continue parsing after reading from the pickled result.
- sqlparser: Parse SQL dumps from postgreSQL and mysql
- timeout: A simple timeout mechanism using SIGALRM
- Tex_CSV_Writer: Write CVS files in a syntax that can be parsed by TeX. Implements same interface as the csv module. Only implements writing.
- trafficshape: Simple traffic shaping configuration: Currently generates Linux tc commands for a traffic shaping config using hfsc for shaping and sfq or red for the leaf qdiscs. We also provide a translator from iptables mangle rules (for traffic marking) to tc commands for filtering. This is needed if we want to do inbound shaping. The Linux kernel doesn't support inbound shaping, so we need to redirect all traffic to an ifb device. On that device we do outbound shaping. Unfortunately when sending traffic to an ifb device with a tc filter + action mirred command this occurs before the PREROUTING chain. This means packets are not yet marked. So we provide a parser to translate mangle rules in the PREROUTING chain to appropriate tc commands. For implementing this we mark packets (using the tc action xt (formerly known as ipt used for iptables marking, a backward-compatibility option can be used) in the original device and then use the marked information in the ifb device to send the traffic to appropriate qdiscs. This supports a two-level approach: Rules in iptables that match on a packet mark (using --mark) are translated to tc commands in the ifb device (they depend on the mark already present). All other rules are translated to a tc action xt (using -j MARK) and a tc action mirred redirect (for sending the traffic to the ifb device) for the original device (e.g. eth0 redirecting to ifb0). The PREROUTING commands by default are directly taken from the running kernel by default (using "iptables -t mangle -S -v")
Download the source at https://sourceforge.net/projects/rsclib/ and install using the standard python setup, e.g.:
python setup.py install --prefix=/usr/local
Version 0.38: Minor feature enhancements
Fix boolean conversion of IP6_Address (and IP4_Address).
- IP6_Address would throw an error when trying a truth-test. Add __nonzero__ (which always returns True even for the 0 Address)
Version 0.37: Minor feature enhancements
Change sort-order of IP_Address, make IP_Address immutable, use metaclass magic to allow copy-constructor semantics.
- Sort order of IP_Address objects (both v4 and v6) now reverses the order of the netmask: If the IP-Address part of the objects to compare are the same, we used to sort by inverse netmask (putting smaller networks with higher netmask first). We now reversed this to be compatible with PostgrSQL cidr type objects.
- All attributes of IP_Address objects are now implemented as properties to return the '_' variant of the attribute. Thus IP_Address objects are (when using the public interface) immutable. Since we already had a __hash__ method this effectively fixes the interface to not allow mutation of objects that are in a dictionary.
- Allow calling the IP_Address constructors with another IP_Address object. Since IP_Address objects are now immutable we use metaclass trickery to return the passed object itself (instead of generating a copy).
Version 0.36: Minor feature enhancements
Allow auto-coercion of comparison parameters. Add parent property and is_sibling test.
- Now comparison operators and 'in' do auto coercion.
- Add parent property (next bigger network)
- Add is_sibling test (same parent)
Version 0.35: Minor feature enhancements
Add 'mask_len' as an alias of 'mask' to IP_Address.
- Need the network mask length (aka prefix length) sometimes as mask_len (e.g. for FFM on github).
Version 0.34: Minor feature enhancements
Fix trafficshape to use new tc syntax. Add label to hexdump.
- The tc command has renamed the ipt action to xt (Linux introduced xtables as a refactoring of iptables), the old ipt is still available in iproute2 but we make xt the default now. A backward-compatibility parameter can be used to get the old behavior.
- Add save-mark to iptables action parser.
- The hexdump class now can generate labels.
Version 0.33: Minor feature enhancements
More fixes for ast_call.
- Add parser for events from asterisk wireshark trace
- Add fail.log for 'real' test
- Don't double-register call with Call_Manager
- Allow explicit matching by account-code
Version 0.32: Minor feature enhancements
More fixes for ast_call.
- Regression test with pyst asterisk emulator
- Fix case where OriginateResponse immediately returns Failure
- Tests for cases where Hangup comes before or after the OriginateResponse
Version 0.31: Minor feature enhancements
Fix ast_call for immediately failing calls. Fix dahdi channel computation in isdn.py.
- Fix OriginateResponse handling in ast_call
- Fix dahdi channel computation, can't directly use the span, use the basechan attribute
Version 0.30: Minor feature enhancements
Fix how dahdi vs. mISDN interpret what is called an interface and what is called a port. In mISDN we can combine several ports (physical lines) to an interface. In dahdi both are the same (a port is a span in dahdi).
- Remove parsing of B- and C- channels from dahdi isdn parser
Version 0.29: Minor feature enhancements
The lcr module is now named isdn. It can now handle isdn interfaces managed by Asterisk DAHDI in addition to Linux Call Router (LCR).
Version 0.28: Minor feature enhancements
Fix inductance formula of Robert Weaver, thanks Robert for pointing me to the correction you did on your new site! For most doctests in the inductance module the error was in the lower percentage points. Add an xxrange iterator to the iter_recipes that can replace pythons native xrange iterator but works with long integers. Needed for some operations on IPv6 addresses in the IP_Address module.
- Fix inductance calculation according to patch from Robert Weaver
- Add xxrange iterator to iter_recipes
- Use new xxrange instead of xrange in IP_Address module, add a test that failed with large numbers for IPv6
Version 0.27: Minor feature enhancements
Add pageurl and pageinfo attributes to HTML_Parser.Page_Tree, other enhancements to HTML_Parser. Add pickle support to parser classes. Fix comparison of IP_Address classes.
- Add pageurl and pageinfo attributes to HTML_Parser.Page_Tree storing information retrieved via geturl and info calls from urllib2.
- Parser classes in stateparser.py and HTML_Parse.py where not pickleable, fixed by removing parser-specific attributes when calling pickle. Note that the parsing cannot be continued after reading class from a pickle.
- Add base_pickler module to allow pickle exceptions
- HTML_Parse: Make Parse_Error a ValueError
- HTML_Parse: Raise line number with exception
- HTML_Parse: Add a timeout
- HTML_Parse: raise Retries_Exceeded with url
- HTML_Parse: url parameter may now be None, not joined with site parameter
- Add pageurl and pageinfo to HTML_Parse
- IP_Address: Fix comparison
- Slight refactoring of NMAP_Parser class
Version 0.26: Minor feature enhancements
Fix double-utf-8-encoding option for sqlparser. Enhance stderr handling for exec_pipe.
- More detected broken encodings for fix_double_encode option
- execute.py: add error message from executed command to message raised by exec_pipe, make stderr output available in non-failing case.
Version 0.25: Minor feature enhancements
Add sqlparser for parsing SQL dumps of PostgreSQL and mysql, add Phone to parse phone numbers.
- sqlparser added
- Phone added for parsing phone numbers
Version 0.24: Minor feature enhancements
Better syntax checks and comparison operators for IP_Address, bug fixes for parser and __str__ for IP_Address.IP6_Address
- IP_Address better syntax checks
- IP_Address __cmp__ and __eq__ improved for comparison with other types
- more regression tests for IP6_Address
- bug fixes in __str__ and parser of IP6_Address
- support for strict checking of netmask (all bits at right of netmask must be zero if strict_mask is True)
Version 0.23: Minor feature enhancements
IP4_Address can now be put in a dict, add a subnets iterator for IP4_Address. Factor IP_Address and add IP6_Address
- Add __hash__ for IP_Address
- The new subnets iterator for IP_Address iterates over all IPs in a subnet. Optionally a netmask can be specified.
- Support for IPv6 addresses
- rename IP4_Address to IP_Address
Version 0.22: Minor feature enhancements
Allow unicode ip address input, hopefully make rsclib installable via pip.
- Address given to IP4_Address constructor now may be unicode
- Add download_url to setup.py to make installable via pip
Version 0.21: Minor feature enhancements
Fix autosuper: allow to inherit from non-autosuper classes, some small fixes to ast_call and lcr parser. Fix ETree pretty-printing. Update Freshmeat to new hostname. Add dotted netmaks parsing to IP4_Address.
- Since python2.6 constructor of "object" do not allow parameters, so we need to strip these when doing the upcall from autosuper. This fails when e.g. inheriting from a non-autosuper enabled class, e.g., class (With_Autosuper, dict) in that case dict would get empty parameters. New implementation finds out if our upcall is to "object", only in that case strip parameters.
- ast_call now processes all queued unhandled events when a call is matched.
- update regression test for ast_call.Call
- lcr parser: fix regex, port can have an empty name.
- Optimize call matching in ast_call: mark call as closed once we are sure about the uniqueid. Add matching of Account-Code.
- Fix ETree pretty-printing: don't print unicode strings when arguments are already converted
- freshmeat.net now is freecode.com (and the API redirects there), update Freshmeat.py to new hostname (including .netrc credentials with compatibility for old name).
- explicit mask paramter of IP4_Address can now be a dotted netmask.
Version 0.20: Not announced on freshmeat
Database value output for ast_cdr, added inductance calculation.
- ast_cdr: Add methods for database values of CDR records -- database values of CDRs are different, they don't include start, end, answer time-stamps but instead only a calldate, in addition the amaflags are numerical in the database.
- added inductance calculation
Version 0.19: Not announced on freshmeat
Extend ETree with a walk method and implement small ldap library
- ETree: add walk method to walk the tree and call an optional pre- and post-hook function
- ldap_lib: common ldap tasks for user and group search, and an iterator for paged search (used with active directory).
Version 0.18: Not announced on freshmeat
Bugfix of ast_call and update for asterisk 1.6, small extension to IP4_Address.
- ast_call: match calls via (unique) account code
- ast_call: State vs ChannelState parameter in Newstate event
- ast_call: handle immediate error from asterisk (e.g. Permission Denied)
- IP4_Address: add netblk (start and end address for address with netmask)
Version 0.17: Not announced on freshmeat
Factor ETree (extended ElementTree) from HTML_Parse. New Freshmeat module to get project information and submit new releases via the new freshmeat REST API. New simple hexdump module.
- New ETree.py (extended ElementTree)
- New Freshmeat.py
- New hexdump.py
- adapt lcr module to new version of Linux Call Router
Version 0.16: Not announced on freshmeat
Add an iptables to tc translator for translating mangle rules in the iptables PREROUTING chain to appropriate tc commands (using an ipt action and mirred redirect actions).
- Add iptables to tc translator to trafficshape.py
Version 0.15: Not announced on freshmeat
Add a framework for traffic shaping with linux iproute (tc). Minor updates to iter_recipes.
- Initial implementation of trafficshape.py
- Add iter_recipes.combinations from python2.6 manpage of itertools for backward compatibility
Version 0.14: Not announced on freshmeat
Add a framework for process pipeline execution, processes can either be python methods or external programs (with parameter list). They can be connected in a pipe and there may be T-points in the pipe, where the pipe forks into two or more pipelines fed by the output of one process.
- Add process pipeline framework
- HTML_Parse now has an explicit translate hook for preprocessing the html page before parsing it. This defaults to the old behaviour of filtering out common characters in broken HTML.
- Add nmap parser (e.g. to generate TeX tables from an nmap scan)
- Fix Lock_Mixin in execute module to remove lockfile at exit, this used to rely on __del__ which breaks in certain cases.
- add file upload to HTML_Parse
Version 0.13: Not announced on freshmeat
Bug-Fix Release: Fix signal handler in timeout.py
- fix signal handler timeout.py
Version 0.12: Not announced on freshmeat
Add a simple timeout mechanism using SIGALRM.
- add timeout.py
Version 0.11: Not announced on freshmeat
Add a parser for CDR records in asterisk. We currently use text-files only, this might be later extended for database use. Some fixes for ast_call, make call-handling more robust (some race conditions would identify events of other calls as belonging to our initiated call). Add an execute module for executing commands and handling IO, this also has a Lock and a Log mixin. Add ocf.py, a small framework for writing OCF scripts for the heartbeat cluster resource manager (probably also usable for the newer version called pacemaker but for now only tested with the version in Debian stable aka lenny). Add lcr.py to model the status of Linux Call Router ISDN lines.
- add ast_cdr.py
- fix ast_call.py
- fix up-chaining in stateparser.py
- add execute.py
- add ocf.py
- add lcr.py
Version 0.10: Not announced on freshmeat
add ast_call for asterisk auto-dialling, small fixes to IP4_Address, add bero*fos configurator, experimental code for checking PDF signature
- add ast_call.py
- Firstname: don't look up names with len < 2
- IP4_Address: some aliases for common functions
- IP4_Address: add __cmp__
- bero.py: bero*fos configurator
- HTML_Parse updated for python 2.5
- stateparser update: use self.matrix by default
- PDF_Signature: experimental code for checking signatures on adobe PDF documents
- iter_recipes: some magic with iterators
Version 0.9: Not announced on freshmeat
Add binom to the Math package, add Firstname, Bug-Fix Release Rational
- binom (n, m) computes the binomial coefficient of n, m.
- Firstname: check if candidate is a first name candidate according to popular web site.
- Rational: On division we could get a negative denominator -- fixed
- make Config_File a descendent of autosuper
Version 0.8: Not announced on freshmeat
Added more documentation. State-machine parser stateparser implemented. Rational number arithmetic package added.
- stateparser implemented (simple state-machine line-oriented configurable parser)
- usage-example of IP4_Address prints debian /etc/network/interfaces entry.
- Math added (Euclids algorithm, gcd, lcm)
- Rational number arithmetics
- cookie processing for HTML_Parse
- basic HTML auth for HTML_Parse
- HTML_Parse: move to urllib2
Version 0.7: Not announced on freshmeat
Small Python library with various things such as Configuration file parsing (in Python syntax), HTML and PDF parsing.
- First Release version