Re: How to know the set of (source?) packages having bytecode executables?

[David MENTRE <dmentre@linux-france.org>]

Hello,

David MENTRE <dmentre@linux-france.org> writes:

> I'll think I'll try a more brute-force approach: installing all
> OCaml-related packages and look for OCaml bytecode binaries in /usr/bin/
> (using the same pattern as proposed in the Ubuntu bug report).
[...]
> After looking at Stephane's gen-binNMU-request.py script, using rmadison
> I should be able to make such a script.

After using Stephane's script as template, I have my Python script to
list all binary packages. I have also some Python code to test if a
file is an OCaml bytecode (match Caml1999X[0-9][0-9][0-9] pattern at the
last 12 bytes of the file).

With "dpkg -S", I should be able to find the relevant binary packages
that contain a bytecode. Is it possible to do the same as "dpkg -S" from
Python code using, I suppose, python-debian code?

Here are my scripts, if it can help others.

Yours,
d.

PS : The ubuntu-hardy-ocaml-bin-pkg.py is Ubuntu Hardy specific but it
     should be easy to adapt it to various Debian/Ubuntu releases.
-- 
GPG/PGP key: A3AD7A2A David MENTRE <dmentre@linux-france.org>
 5996 CC46 4612 9CA4 3562  D7AC 6C67 9E96 A3AD 7A2A

#!/usr/bin/python
# -*- coding: utf-8 -*-

# Extracted from gen-binNMU-request.py:
#  * download()
# Use of rmadison also extracted from this script. Thanks Stephane Glondu!

import sys, os, re
from commands import getoutput, getstatusoutput
from httplib import HTTPConnection

reference_arch = "i386"
madison_split_regexp = re.compile("[ |,\n]+")

def download(site, file_path):
    c = HTTPConnection(site)
    c.request("GET", "/%s" % file_path)
    r = c.getresponse().read()
    c.close()
    return r

def get_binary_packages(src_pkg):
    if src_pkg == "":
        return []
    rmadison_raw = getoutput("rmadison -u ubuntu -s hardy -S %s" % src_pkg)
    binary_packages = []
    for line in filter(None, rmadison_raw.split("\n")):
        splitted = madison_split_regexp.split(line.strip())
        binary_package_name = splitted[0]
        version = splitted[1]
        available_archs = splitted[3:]
        #print binary_package_name, version, available_archs
        if "all" in available_archs:
            binary_packages.append(binary_package_name)
        elif reference_arch in available_archs:
            binary_packages.append(binary_package_name)
    #print "Kept: ", binary_packages
    return binary_packages

def output_aptget_command(package_list):
    print "## To install ##"
    print "apt-get \\"
    for pkg in package_list[:-1]:
        print "   ", pkg, "\\"
    print "   ", package_list[-1]

def main():
    ocaml_sources_packages = download("pkg-ocaml-maint.alioth.debian.org",
                                      "ocaml_src_pkgs.txt").split("\n")
    all_binary_packages = []
    for src_pkg in ocaml_sources_packages:
        all_binary_packages.extend(get_binary_packages(src_pkg))
    output_aptget_command(all_binary_packages)

main()

#!/usr/bin/python

import sys
import os
import re

def is_ocaml_bytecode(filename):
    f = open(filename, "r")
    f.seek(-12, 2) # go to 12 bytes before the end of file
    bytes = f.read(12)
    f.close()
    if re.match('Caml1999X[0-9][0-9][0-9]', bytes):
        return True
    else:
        return False

for filename in sys.argv[1:]:
    print filename, is_ocaml_bytecode(filename)