3 # Various different sanity checks
4 # Copyright (C) 2000, 2001, 2002, 2003 James Troup <james@nocrew.org>
5 # $Id: tea,v 1.24 2003-09-24 00:13:46 troup Exp $
7 # This program is free software; you can redistribute it and/or modify
8 # it under the terms of the GNU General Public License as published by
9 # the Free Software Foundation; either version 2 of the License, or
10 # (at your option) any later version.
12 # This program is distributed in the hope that it will be useful,
13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 # GNU General Public License for more details.
17 # You should have received a copy of the GNU General Public License
18 # along with this program; if not, write to the Free Software
19 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
21 ################################################################################
23 # And, lo, a great and menacing voice rose from the depths, and with
24 # great wrath and vehemence it's voice boomed across the
25 # land... ``hehehehehehe... that *tickles*''
28 ################################################################################
30 import commands, os, pg, stat, string, sys, tempfile, time;
31 import db_access, utils;
32 import apt_pkg, apt_inst;
34 ################################################################################
43 current_time = time.time();
45 ################################################################################
47 def usage(exit_code=0):
48 print """Usage: tea MODE
49 Run various sanity checks of the archive and/or database.
51 -h, --help show this help and exit.
53 The following MODEs are available:
55 md5sums - validate the md5sums stored in the database
56 files - check files in the database against what's in the archive
57 dsc-syntax - validate the syntax of .dsc files in the archive
58 missing-overrides - check for missing overrides
59 source-in-one-dir - ensure the source for each package is in one directory
60 timestamps - check for future timestamps in .deb's
61 tar-gz-in-dsc - ensure each .dsc lists a .tar.gz file
62 validate-indices - ensure files mentioned in Packages & Sources exist
63 files-not-symlinks - check files in the database aren't symlinks
67 ################################################################################
69 def process_dir (unused, dirname, filenames):
70 global waste, db_files, excluded;
72 if dirname.find('/disks-') != -1 or dirname.find('upgrade-') != -1:
74 # hack; can't handle .changes files
75 if dirname.find('proposed-updates') != -1:
77 for name in filenames:
78 filename = os.path.abspath(dirname+'/'+name);
79 filename = filename.replace('potato-proposed-updates', 'proposed-updates');
80 if os.path.isfile(filename) and not os.path.islink(filename) and not db_files.has_key(filename) and not excluded.has_key(filename):
81 waste += os.stat(filename)[stat.ST_SIZE];
84 ################################################################################
89 print "Building list of database files...";
90 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id")
95 filename = os.path.abspath(i[0] + i[1]);
96 db_files[filename] = "";
97 if os.access(filename, os.R_OK) == 0:
98 utils.warn("'%s' doesn't exist." % (filename));
100 file = utils.open_file(Cnf["Dir::Override"]+'override.unreferenced');
101 for filename in file.readlines():
102 filename = filename[:-1];
103 excluded[filename] = "";
105 print "Checking against existent files...";
107 os.path.walk(Cnf["Dir::Root"]+'dists/', process_dir, None);
110 print "%s wasted..." % (utils.size_type(waste));
112 ################################################################################
117 for component in Cnf.SubTree("Component").List():
118 if component == "mixed":
120 component = component.lower();
121 list_filename = '%s%s_%s_source.list' % (Cnf["Dir::Lists"], suite, component);
122 list_file = utils.open_file(list_filename);
123 for line in list_file.readlines():
126 utils.parse_changes(file, dsc_whitespace_rules=1);
127 except utils.invalid_dsc_format_exc, line:
128 utils.warn("syntax error in .dsc file '%s', line %s." % (file, line));
132 utils.warn("Found %s invalid .dsc files." % (count));
134 ################################################################################
136 def check_override():
137 for suite in [ "stable", "unstable" ]:
139 print "-------------"
141 suite_id = db_access.get_suite_id(suite);
142 q = projectB.query("""
143 SELECT DISTINCT b.package FROM binaries b, bin_associations ba
144 WHERE b.id = ba.bin AND ba.suite = %s AND NOT EXISTS
145 (SELECT 1 FROM override o WHERE o.suite = %s AND o.package = b.package)"""
146 % (suite_id, suite_id));
148 q = projectB.query("""
149 SELECT DISTINCT s.source FROM source s, src_associations sa
150 WHERE s.id = sa.source AND sa.suite = %s AND NOT EXISTS
151 (SELECT 1 FROM override o WHERE o.suite = %s and o.package = s.source)"""
152 % (suite_id, suite_id));
155 ################################################################################
157 # Ensure that the source files for any given package is all in one
158 # directory so that 'apt-get source' works...
160 def check_source_in_one_dir():
161 # Not the most enterprising method, but hey...
163 q = projectB.query("SELECT id FROM source;");
164 for i in q.getresult():
166 q2 = projectB.query("""
167 SELECT l.path, f.filename FROM files f, dsc_files df, location l WHERE df.source = %s AND f.id = df.file AND l.id = f.location"""
172 for j in q2.getresult():
173 filename = j[0] + j[1];
174 path = os.path.dirname(filename);
177 first_filename = filename;
178 elif first_path != path:
179 symlink = path + '/' + os.path.basename(first_filename);
180 if not os.path.exists(symlink):
182 print "WOAH, we got a live one here... %s [%s] {%s}" % (filename, source_id, symlink);
185 print "Found %d source packages where the source is not all in one directory." % (broken_count);
187 ################################################################################
190 print "Getting file information from database...";
191 q = projectB.query("SELECT l.path, f.filename, f.md5sum, f.size FROM files f, location l WHERE f.location = l.id")
194 print "Checking file md5sums & sizes...";
196 filename = os.path.abspath(i[0] + i[1]);
200 file = utils.open_file(filename);
202 utils.warn("can't open '%s'." % (filename));
204 md5sum = apt_pkg.md5sum(file);
205 size = os.stat(filename)[stat.ST_SIZE];
206 if md5sum != db_md5sum:
207 utils.warn("**WARNING** md5sum mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, md5sum, db_md5sum));
209 utils.warn("**WARNING** size mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, size, db_size));
213 ################################################################################
215 # Check all files for timestamps in the future; common from hardware
216 # (e.g. alpha) which have far-future dates as their default dates.
218 def Ent(Kind,Name,Link,Mode,UID,GID,Size,MTime,Major,Minor):
221 if MTime > current_time:
222 future_files[current_file] = MTime;
223 print "%s: %s '%s','%s',%u,%u,%u,%u,%u,%u,%u" % (current_file, Kind,Name,Link,Mode,UID,GID,Size, MTime, Major, Minor);
225 def check_timestamps():
228 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id AND f.filename ~ '.deb$'")
233 filename = os.path.abspath(i[0] + i[1]);
234 if os.access(filename, os.R_OK):
235 file = utils.open_file(filename);
236 current_file = filename;
237 sys.stderr.write("Processing %s.\n" % (filename));
238 apt_inst.debExtract(file,Ent,"control.tar.gz");
240 apt_inst.debExtract(file,Ent,"data.tar.gz");
242 print "Checked %d files (out of %d)." % (count, len(db_files.keys()));
244 ################################################################################
246 def check_missing_tar_gz_in_dsc():
249 print "Building list of database files...";
250 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id AND f.filename ~ '.dsc$'");
253 print "Checking %d files..." % len(ql);
255 print "No files to check."
257 filename = os.path.abspath(i[0] + i[1]);
259 # NB: don't enforce .dsc syntax
260 dsc = utils.parse_changes(filename);
262 utils.fubar("error parsing .dsc file '%s'." % (filename));
263 dsc_files = utils.build_file_list(dsc, is_a_dsc=1);
265 for file in dsc_files.keys():
266 m = utils.re_issource.match(file);
268 utils.fubar("%s not recognised as source." % (file));
270 if type == "orig.tar.gz" or type == "tar.gz":
273 utils.warn("%s has no .tar.gz in the .dsc file." % (file));
277 utils.warn("Found %s invalid .dsc files." % (count));
280 ################################################################################
282 def validate_sources(suite, component):
283 filename = "%s/dists/%s/%s/source/Sources.gz" % (Cnf["Dir::Root"], suite, component);
284 print "Processing %s..." % (filename);
285 # apt_pkg.ParseTagFile needs a real file handle and can't handle a GzipFile instance...
286 temp_filename = tempfile.mktemp();
287 fd = os.open(temp_filename, os.O_RDWR|os.O_CREAT|os.O_EXCL, 0700);
289 (result, output) = commands.getstatusoutput("gunzip -c %s > %s" % (filename, temp_filename));
291 sys.stderr.write("Gunzip invocation failed!\n%s\n" % (output));
293 sources = utils.open_file(temp_filename);
294 Sources = apt_pkg.ParseTagFile(sources);
295 while Sources.Step():
296 source = Sources.Section.Find('Package');
297 directory = Sources.Section.Find('Directory');
298 files = Sources.Section.Find('Files');
299 for i in files.split('\n'):
301 (md5, size, name) = s;
302 filename = "%s/%s/%s" % (Cnf["Dir::Root"], directory, name);
303 if not os.path.exists(filename):
304 if directory.find("potato") == -1:
305 print "W: %s missing." % (filename);
307 pool_location = utils.poolify (source, component);
308 pool_filename = "%s/%s/%s" % (Cnf["Dir::Pool"], pool_location, name);
309 if not os.path.exists(pool_filename):
310 print "E: %s missing (%s)." % (filename, pool_filename);
313 pool_filename = os.path.normpath(pool_filename);
314 filename = os.path.normpath(filename);
315 src = utils.clean_symlink(pool_filename, filename, Cnf["Dir::Root"]);
316 print "Symlinking: %s -> %s" % (filename, src);
317 #os.symlink(src, filename);
318 os.unlink(temp_filename);
320 ########################################
322 def validate_packages(suite, component, architecture):
323 filename = "%s/dists/%s/%s/binary-%s/Packages" \
324 % (Cnf["Dir::Root"], suite, component, architecture);
325 print "Processing %s..." % (filename);
326 packages = utils.open_file(filename);
327 Packages = apt_pkg.ParseTagFile(packages);
328 while Packages.Step():
329 filename = "%s/%s" % (Cnf["Dir::Root"], Packages.Section.Find('Filename'));
330 if not os.path.exists(filename):
331 print "W: %s missing." % (filename);
334 ########################################
336 def check_indices_files_exist():
337 for suite in [ "stable", "testing", "unstable" ]:
338 for component in Cnf.ValueList("Suite::%s::Components" % (suite)):
339 architectures = Cnf.ValueList("Suite::%s::Architectures" % (suite));
340 for arch in map(string.lower, architectures):
342 validate_sources(suite, component);
346 validate_packages(suite, component, arch);
348 ################################################################################
350 def check_files_not_symlinks():
351 print "Building list of database files... ",;
352 before = time.time();
353 q = projectB.query("SELECT l.path, f.filename, f.id FROM files f, location l WHERE f.location = l.id")
354 print "done. (%d seconds)" % (int(time.time()-before));
355 q_files = q.getresult();
358 # q = projectB.query("SELECT l.path, c.name, l.id FROM location l, component c WHERE l.component = c.id");
359 # for i in q.getresult():
360 # path = os.path.normpath(i[0] + i[1]);
361 # locations[path] = (i[0], i[2]);
363 # q = projectB.query("BEGIN WORK");
365 filename = os.path.normpath(i[0] + i[1]);
367 if os.access(filename, os.R_OK) == 0:
368 utils.warn("%s: doesn't exist." % (filename));
370 if os.path.islink(filename):
371 utils.warn("%s: is a symlink." % (filename));
372 # You probably don't want to use the rest of this...
373 # print "%s: is a symlink." % (filename);
374 # dest = os.readlink(filename);
375 # if not os.path.isabs(dest):
376 # dest = os.path.normpath(os.path.join(os.path.dirname(filename), dest));
377 # print "--> %s" % (dest);
378 # # Determine suitable location ID
379 # # [in what must be the suckiest way possible?]
380 # location_id = None;
381 # for path in locations.keys():
382 # if dest.find(path) == 0:
383 # (location, location_id) = locations[path];
385 # if not location_id:
386 # utils.fubar("Can't find location for %s (%s)." % (dest, filename));
387 # new_filename = dest.replace(location, "");
388 # q = projectB.query("UPDATE files SET filename = '%s', location = %s WHERE id = %s" % (new_filename, location_id, file_id));
389 # q = projectB.query("COMMIT WORK");
391 ################################################################################
394 global Cnf, projectB, db_files, waste, excluded;
396 Cnf = utils.get_conf();
397 Arguments = [('h',"help","Tea::Options::Help")];
399 if not Cnf.has_key("Tea::Options::%s" % (i)):
400 Cnf["Tea::Options::%s" % (i)] = "";
402 args = apt_pkg.ParseCommandLine(Cnf, Arguments, sys.argv);
404 Options = Cnf.SubTree("Tea::Options")
409 utils.warn("tea requires at least one argument");
412 utils.warn("tea accepts only one argument");
414 mode = args[0].lower();
416 projectB = pg.connect(Cnf["DB::Name"], Cnf["DB::Host"], int(Cnf["DB::Port"]));
417 db_access.init(Cnf, projectB);
419 if mode == "md5sums":
421 elif mode == "files":
423 elif mode == "dsc-syntax":
425 elif mode == "missing-overrides":
427 elif mode == "source-in-one-dir":
428 check_source_in_one_dir();
429 elif mode == "timestamps":
431 elif mode == "tar-gz-in-dsc":
432 check_missing_tar_gz_in_dsc();
433 elif mode == "validate-indices":
434 check_indices_files_exist();
435 elif mode == "files-not-symlinks":
436 check_files_not_symlinks();
438 utils.warn("unknown mode '%s'" % (mode));
441 ################################################################################
443 if __name__ == '__main__':