3 Create all the contents files
5 @contact: Debian FTPMaster <ftpmaster@debian.org>
6 @copyright: 2008, 2009 Michael Casadevall <mcasadevall@debian.org>
7 @copyright: 2009 Mike O'Connor <stew@debian.org>
8 @license: GNU General Public License version 2 or later
11 ################################################################################
13 # This program is free software; you can redistribute it and/or modify
14 # it under the terms of the GNU General Public License as published by
15 # the Free Software Foundation; either version 2 of the License, or
16 # (at your option) any later version.
18 # This program is distributed in the hope that it will be useful,
19 # but WITHOUT ANY WARRANTY; without even the implied warranty of
20 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
21 # GNU General Public License for more details.
23 # You should have received a copy of the GNU General Public License
24 # along with this program; if not, write to the Free Software
25 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
27 ################################################################################
29 # <Ganneff> there is the idea to slowly replace contents files
30 # <Ganneff> with a new generation of such files.
31 # <Ganneff> having more info.
33 # <Ganneff> of course that wont help for now where we need to generate them :)
35 ################################################################################
44 from daklib import utils
45 from daklib.config import Config
46 from daklib.dbconn import DBConn
47 ################################################################################
49 def usage (exit_code=0):
50 print """Usage: dak contents [options] command [arguments]
54 generate Contents-$arch.gz files
57 scan the debs in the existing pool and load contents in the the database
60 remove files/paths which are no longer referenced by a binary
64 show this help and exit
67 show verbose information messages
70 supress all output but errors
72 -s, --suite={stable,testing,unstable,...}
73 only operate on a single suite
75 -a, --arch={i386,amd64}
76 only operate on a single architecture
80 ################################################################################
82 # where in dak.conf all of our configuration will be stowed
84 options_prefix = "Contents"
85 options_prefix = "%s::Options" % options_prefix
87 log = logging.getLogger()
89 ################################################################################
91 # we unfortunately still have broken stuff in headers
92 latin1_q = """SET CLIENT_ENCODING TO 'LATIN1'"""
94 # get all the arches delivered for a given suite
95 # this should probably exist somehere common
96 arches_q = """PREPARE arches_q as
97 SELECT s.architecture, a.arch_string
98 FROM suite_architectures s
99 JOIN architecture a ON (s.architecture=a.id)
102 # find me the .deb for a given binary id
103 debs_q = """PREPARE debs_q as
104 SELECT b.id, f.filename FROM bin_assoc_by_arch baa
105 JOIN binaries b ON baa.bin=b.id
106 JOIN files f ON b.file=f.id
110 # ask if we already have contents associated with this binary
111 olddeb_q = """PREPARE olddeb_q as
112 SELECT 1 FROM content_associations
113 WHERE binary_pkg = $1
116 # find me all of the contents for a given .deb
117 contents_q = """PREPARE contents_q as
118 SELECT (p.path||'/'||n.file) AS fn,
119 comma_separated_list(s.section||'/'||b.package)
120 FROM content_associations c
121 JOIN content_file_paths p ON (c.filepath=p.id)
122 JOIN content_file_names n ON (c.filename=n.id)
123 JOIN binaries b ON (b.id=c.binary_pkg)
124 JOIN bin_associations ba ON (b.id=ba.bin)
125 JOIN override o ON (o.package=b.package)
126 JOIN section s ON (s.id=o.section)
127 WHERE (b.architecture = $1 OR b.architecture = $2)
135 # find me all of the contents for a given .udeb
136 udeb_contents_q = """PREPARE udeb_contents_q as
137 SELECT (p.path||'/'||n.file) as fn,
138 comma_separated_list(s.section||'/'||b.package)
139 FROM content_associations c
140 JOIN content_file_paths p ON (c.filepath=p.id)
141 JOIN content_file_names n ON (c.filename=n.id)
142 JOIN binaries b ON (b.id=c.binary_pkg)
143 JOIN bin_associations ba ON (b.id=ba.bin)
144 JOIN override o ON (o.package=b.package)
145 JOIN section s ON (s.id=o.section)
154 # clear out all of the temporarily stored content associations
155 # this should be run only after p-a has run. after a p-a
156 # run we should have either accepted or rejected every package
157 # so there should no longer be anything in the queue
158 remove_pending_contents_cruft_q = """DELETE FROM pending_content_associations"""
160 # delete any filenames we are storing which have no binary associated with them
161 remove_filename_cruft_q = """DELETE FROM content_file_names
162 WHERE id IN (SELECT cfn.id FROM content_file_names cfn
163 LEFT JOIN content_associations ca
164 ON ca.filename=cfn.id
165 WHERE ca.id IS NULL)""" );
167 # delete any paths we are storing which have no binary associated with them
168 remove_filepath_cruft_q = """DELETE FROM content_file_paths
169 WHERE id IN (SELECT cfn.id FROM content_file_paths cfn
170 LEFT JOIN content_associations ca
171 ON ca.filepath=cfn.id
172 WHERE ca.id IS NULL)"""
173 class Contents(object):
175 Class capable of generating Contents-$arch.gz files
177 Usage GenerateContents().generateContents( ["main","contrib","non-free"] )
183 def _getHeader(self):
185 Internal method to return the header for Contents.gz files
187 This is boilerplate which explains the contents of the file and how
190 if self.header == None:
191 if Config().has_key("Contents::Header"):
193 h = open(os.path.join( Config()["Dir::Templates"],
194 Config()["Contents::Header"] ), "r")
195 self.header = h.read()
196 print( "header: %s" % self.header )
199 log.error( "error opening header file: %d\n%s" % (Config()["Contents::Header"],
200 traceback.format_exc() ))
208 # goal column for section column
211 def _write_content_file(self, cursor, filename):
213 Internal method for writing all the results to a given file.
214 The cursor should have a result set generated from a query already.
216 f = gzip.open(Config()["Dir::Root"] + filename, "w")
218 header = self._getHeader()
224 contents = cursor.fetchone()
229 int( math.ceil( (self._goal_column - len(contents[0])) / 8) ) )
230 f.write(contents[0] + ( '\t' * num_tabs ) + contents[-1] + "\n")
237 remove files/paths from the DB which are no longer referenced
238 by binaries and clean the temporary table
240 cursor = DBConn().cursor();
241 cursor.execute( "BEGIN WORK" )
242 cursor.execute( remove_pending_contents_cruft_q )
243 cursor.execute( remove_filename_cruft_q )
244 cursor.execute( remove_filepath_cruft_q )
245 cursor.execute( "COMMIT" )
250 scan the existing debs in the pool to populate the contents database tables
252 pooldir = Config()[ 'Dir::Pool' ]
254 cursor = DBConn().cursor();
255 cursor.execute( latin1_q )
256 cursor.execute( debs_q )
257 cursor.execute( olddeb_q )
258 cursor.execute( arches_q )
260 suites = self._suites()
261 for suite in [i.lower() for i in suites]:
262 suite_id = DBConn().get_suite_id(suite)
264 arch_list = self._arches(cursor, suite_id)
265 arch_all_id = DBConn().get_architecture_id("all")
266 for arch_id in arch_list:
267 cursor.execute( "EXECUTE debs_q(%d, %d)" % ( suite_id, arch_id[0] ) )
269 debs = cursor.fetchall()
273 cursor.execute( "EXECUTE olddeb_q(%d)" % (deb[0] ) )
274 old = cursor.fetchone()
276 log.debug( "already imported: %s" % deb[1] )
278 debfile = os.path.join( pooldir, deb[1] )
279 if os.path.exists( debfile ):
280 contents = utils.generate_contents_information( debfile )
281 DBConn().insert_content_paths(deb[0], contents)
282 log.info( "imported (%d/%d): %s" % (count,len(debs),deb[1] ) )
284 log.error( "missing .deb: %s" % deb[1] )
288 Generate Contents-$arch.gz files for every aviailable arch in each given suite.
290 cursor = DBConn().cursor();
292 cursor.execute( arches_q )
293 cursor.execute( contents_q )
294 cursor.execute( udeb_contents_q )
296 suites = self._suites()
298 # Get our suites, and the architectures
299 for suite in [i.lower() for i in suites]:
300 suite_id = DBConn().get_suite_id(suite)
301 arch_list = self._arches(cursor, suite_id)
303 arch_all_id = DBConn().get_architecture_id("all")
305 for arch_id in arch_list:
306 cursor.execute( "EXECUTE contents_q(%d,%d,%d,%d)" % (arch_id[0], arch_all_id, suite_id, suite_id))
307 self._write_content_file(cursor, "dists/%s/Contents-%s.gz" % (suite, arch_id[1]))
309 # The MORE fun part. Ok, udebs need their own contents files, udeb, and udeb-nf (not-free)
310 # This is HORRIBLY debian specific :-/
312 section_id = DBConn().get_section_id('debian-installer') # all udebs should be here)
314 cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d)" % (section_id, suite_id, suite_id))
315 self._write_content_file(cursor, "dists/%s/Contents-udeb.gz" % suite)
317 # Once more, with non-free
318 section_id = DBConn().get_section_id('non-free/debian-installer') # all udebs should be here)
321 cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d)" % (section_id, suite_id, suite_id))
322 self._write_content_file(cursor, "dists/%s/Contents-udeb-nf.gz" % suite)
325 ################################################################################
329 return a list of suites to operate on
331 if Config().has_key( "%s::%s" %(options_prefix,"Suite")):
332 suites = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Suite")])
334 suites = Config().SubTree("Suite").List()
338 def _arches(self, cursor, suite):
340 return a list of archs to operate on
343 if Config().has_key( "%s::%s" %(options_prefix,"Arch")):
344 archs = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Arch")])
345 for arch_name in archs:
346 arch_list.append((DBConn().get_architecture_id(arch_name), arch_name))
348 cursor.execute("EXECUTE arches_q(%d)" % (suite))
350 r = cursor.fetchone()
354 if r[1] != "source" and r[1] != "all":
355 arch_list.append((r[0], r[1]))
359 ################################################################################
364 arguments = [('h',"help", "%s::%s" % (options_prefix,"Help")),
365 ('s',"suite", "%s::%s" % (options_prefix,"Suite"),"HasArg"),
366 ('q',"quiet", "%s::%s" % (options_prefix,"Quiet")),
367 ('v',"verbose", "%s::%s" % (options_prefix,"Verbose")),
368 ('a',"arch", "%s::%s" % (options_prefix,"Arch"),"HasArg"),
371 commands = {'generate' : Contents.generate,
372 'bootstrap' : Contents.bootstrap,
373 'cruft' : Contents.cruft,
377 if cnf.has_key("%s::%s" % (options_prefix,"Quiet")):
380 elif cnf.has_key("%s::%s" % (options_prefix,"Verbose")):
384 logging.basicConfig( level=level,
385 format='%(asctime)s %(levelname)s %(message)s',
386 stream = sys.stderr )
388 args = apt_pkg.ParseCommandLine(cnf.Cnf, arguments,sys.argv)
390 if (len(args) < 1) or not commands.has_key(args[0]):
393 if cnf.has_key("%s::%s" % (options_prefix,"Help")):
396 commands[args[0]](Contents())
398 if __name__ == '__main__':