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 ################################################################################
43 from daklib import utils
44 from daklib.binary import Binary
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 # get all the arches delivered for a given suite
92 # this should probably exist somehere common
93 arches_q = """PREPARE arches_q(int) as
94 SELECT s.architecture, a.arch_string
95 FROM suite_architectures s
96 JOIN architecture a ON (s.architecture=a.id)
99 # find me the .deb for a given binary id
100 debs_q = """PREPARE debs_q(int, int) as
101 SELECT b.id, f.filename FROM bin_assoc_by_arch baa
102 JOIN binaries b ON baa.bin=b.id
103 JOIN files f ON b.file=f.id
107 # ask if we already have contents associated with this binary
108 olddeb_q = """PREPARE olddeb_q(int) as
109 SELECT 1 FROM content_associations
110 WHERE binary_pkg = $1
113 # find me all of the contents for a given .deb
114 contents_q = """PREPARE contents_q(int,int,int,int) as
115 SELECT (p.path||'/'||n.file) AS fn,
116 comma_separated_list(s.section||'/'||b.package)
117 FROM content_associations c
118 JOIN content_file_paths p ON (c.filepath=p.id)
119 JOIN content_file_names n ON (c.filename=n.id)
120 JOIN binaries b ON (b.id=c.binary_pkg)
121 JOIN bin_associations ba ON (b.id=ba.bin)
122 JOIN override o ON (o.package=b.package)
123 JOIN section s ON (s.id=o.section)
124 WHERE (b.architecture = $1 OR b.architecture = $2)
132 udeb_contents_q = """PREPARE udeb_contents_q(int,int,int,int,int) as
133 SELECT (p.path||'/'||n.file) as fn,
134 comma_separated_list(s.section||'/'||b.package)
135 FROM content_associations c
136 JOIN content_file_paths p ON (c.filepath=p.id)
137 JOIN content_file_names n ON (c.filename=n.id)
138 JOIN binaries b ON (b.id=c.binary_pkg)
139 JOIN bin_associations ba ON (b.id=ba.bin)
140 JOIN override o ON (o.package=b.package)
141 JOIN section s ON (s.id=o.section)
142 WHERE (b.architecture = $1 OR b.architecture = $2)
152 # clear out all of the temporarily stored content associations
153 # this should be run only after p-a has run. after a p-a
154 # run we should have either accepted or rejected every package
155 # so there should no longer be anything in the queue
156 remove_pending_contents_cruft_q = """DELETE FROM pending_content_associations"""
158 # delete any filenames we are storing which have no binary associated with them
159 remove_filename_cruft_q = """DELETE FROM content_file_names
160 WHERE id IN (SELECT cfn.id FROM content_file_names cfn
161 LEFT JOIN content_associations ca
162 ON ca.filename=cfn.id
163 WHERE ca.id IS NULL)"""
165 # delete any paths we are storing which have no binary associated with them
166 remove_filepath_cruft_q = """DELETE FROM content_file_paths
167 WHERE id IN (SELECT cfn.id FROM content_file_paths cfn
168 LEFT JOIN content_associations ca
169 ON ca.filepath=cfn.id
170 WHERE ca.id IS NULL)"""
171 class Contents(object):
173 Class capable of generating Contents-$arch.gz files
175 Usage GenerateContents().generateContents( ["main","contrib","non-free"] )
181 def reject(self, message):
182 log.error("E: %s" % message)
184 def _getHeader(self):
186 Internal method to return the header for Contents.gz files
188 This is boilerplate which explains the contents of the file and how
191 if self.header == None:
192 if Config().has_key("Contents::Header"):
194 h = open(os.path.join( Config()["Dir::Templates"],
195 Config()["Contents::Header"] ), "r")
196 self.header = h.read()
199 log.error( "error opening header file: %d\n%s" % (Config()["Contents::Header"],
200 traceback.format_exc() ))
207 # goal column for section column
210 def _write_content_file(self, cursor, filename):
212 Internal method for writing all the results to a given file.
213 The cursor should have a result set generated from a query already.
215 filepath = Config()["Contents::Root"] + filename
216 filedir = os.path.dirname(filepath)
217 if not os.path.isdir(filedir):
219 f = gzip.open(filepath, "w")
221 header = self._getHeader()
227 contents = cursor.fetchone()
232 int(math.ceil((self._goal_column - len(contents[0])-1) / 8)))
233 f.write(contents[0] + ( '\t' * num_tabs ) + contents[-1] + "\n")
240 remove files/paths from the DB which are no longer referenced
241 by binaries and clean the temporary table
243 cursor = DBConn().cursor();
244 cursor.execute( "BEGIN WORK" )
245 cursor.execute( remove_pending_contents_cruft_q )
246 cursor.execute( remove_filename_cruft_q )
247 cursor.execute( remove_filepath_cruft_q )
248 cursor.execute( "COMMIT" )
253 scan the existing debs in the pool to populate the contents database tables
255 pooldir = Config()[ 'Dir::Pool' ]
257 cursor = DBConn().cursor();
258 DBConn().prepare("debs_q",debs_q)
259 DBConn().prepare("olddeb_q",olddeb_q)
260 DBConn().prepare("arches_q",arches_q)
262 suites = self._suites()
263 for suite in [i.lower() for i in suites]:
264 suite_id = DBConn().get_suite_id(suite)
266 arch_list = self._arches(cursor, suite_id)
267 arch_all_id = DBConn().get_architecture_id("all")
268 for arch_id in arch_list:
269 cursor.execute( "EXECUTE debs_q(%d, %d)" % ( suite_id, arch_id[0] ) )
273 deb = cursor.fetchone()
277 cursor1 = DBConn().cursor();
278 cursor1.execute( "EXECUTE olddeb_q(%d)" % (deb[0] ) )
279 old = cursor1.fetchone()
281 log.debug( "already imported: %s" % (deb[1]) )
283 log.debug( "scanning: %s" % (deb[1]) )
284 debfile = os.path.join( pooldir, deb[1] )
285 if os.path.exists( debfile ):
286 Binary(debfile, self.reject).scan_package(deb[0],True)
288 log.error("missing .deb: %s" % deb[1])
292 Generate Contents-$arch.gz files for every available arch in each given suite.
294 cursor = DBConn().cursor();
296 DBConn().prepare( "arches_q", arches_q )
297 DBConn().prepare( "contents_q", contents_q )
298 DBConn().prepare( "udeb_contents_q", udeb_contents_q )
300 debtype_id=DBConn().get_override_type_id("deb")
301 udebtype_id=DBConn().get_override_type_id("udeb")
303 suites = self._suites()
305 # Get our suites, and the architectures
306 for suite in [i.lower() for i in suites]:
307 suite_id = DBConn().get_suite_id(suite)
308 arch_list = self._arches(cursor, suite_id)
310 arch_all_id = DBConn().get_architecture_id("all")
312 for arch_id in arch_list:
313 cursor.execute("EXECUTE contents_q(%d,%d,%d,%d)" % (arch_id[0], arch_all_id, suite_id, debtype_id))
314 self._write_content_file(cursor, "dists/%s/Contents-%s.gz" % (suite, arch_id[1]))
316 # The MORE fun part. Ok, udebs need their own contents files, udeb, and udeb-nf (not-free)
317 # This is HORRIBLY debian specific :-/
318 for section, fn_pattern in [("debian-installer","dists/%s/Contents-udeb-%s.gz"),
319 ("non-free/debian-installer", "dists/%s/Contents-udeb-nf-%s.gz")]:
321 for arch_id in arch_list:
322 section_id = DBConn().get_section_id(section) # all udebs should be here)
324 cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d,%d,%d)" % (arch_id[0], arch_all_id, section_id, suite_id, udebtype_id))
326 self._write_content_file(cursor, fn_pattern % (suite, arch_id[1]))
329 ################################################################################
333 return a list of suites to operate on
335 if Config().has_key( "%s::%s" %(options_prefix,"Suite")):
336 suites = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Suite")])
338 suites = Config().SubTree("Suite").List()
342 def _arches(self, cursor, suite):
344 return a list of archs to operate on
347 if Config().has_key( "%s::%s" %(options_prefix,"Arch")):
348 archs = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Arch")])
349 for arch_name in archs:
350 arch_list.append((DBConn().get_architecture_id(arch_name), arch_name))
352 cursor.execute("EXECUTE arches_q(%d)" % (suite))
354 r = cursor.fetchone()
358 if r[1] != "source" and r[1] != "all":
359 arch_list.append((r[0], r[1]))
363 ################################################################################
368 arguments = [('h',"help", "%s::%s" % (options_prefix,"Help")),
369 ('s',"suite", "%s::%s" % (options_prefix,"Suite"),"HasArg"),
370 ('q',"quiet", "%s::%s" % (options_prefix,"Quiet")),
371 ('v',"verbose", "%s::%s" % (options_prefix,"Verbose")),
372 ('a',"arch", "%s::%s" % (options_prefix,"Arch"),"HasArg"),
375 commands = {'generate' : Contents.generate,
376 'bootstrap' : Contents.bootstrap,
377 'cruft' : Contents.cruft,
380 args = apt_pkg.ParseCommandLine(cnf.Cnf, arguments,sys.argv)
382 if (len(args) < 1) or not commands.has_key(args[0]):
385 if cnf.has_key("%s::%s" % (options_prefix,"Help")):
389 if cnf.has_key("%s::%s" % (options_prefix,"Quiet")):
392 elif cnf.has_key("%s::%s" % (options_prefix,"Verbose")):
396 logging.basicConfig( level=level,
397 format='%(asctime)s %(levelname)s %(message)s',
398 stream = sys.stderr )
400 commands[args[0]](Contents())
402 if __name__ == '__main__':