3 # Various different sanity checks
4 # Copyright (C) 2000, 2001, 2002, 2003, 2004 James Troup <james@nocrew.org>
5 # $Id: tea,v 1.28 2004-06-17 14:57:47 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, 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
64 validate-builddeps - validate build-dependencies of .dsc files in the archive
68 ################################################################################
70 def process_dir (unused, dirname, filenames):
71 global waste, db_files, excluded;
73 if dirname.find('/disks-') != -1 or dirname.find('upgrade-') != -1:
75 # hack; can't handle .changes files
76 if dirname.find('proposed-updates') != -1:
78 for name in filenames:
79 filename = os.path.abspath(dirname+'/'+name);
80 filename = filename.replace('potato-proposed-updates', 'proposed-updates');
81 if os.path.isfile(filename) and not os.path.islink(filename) and not db_files.has_key(filename) and not excluded.has_key(filename):
82 waste += os.stat(filename)[stat.ST_SIZE];
85 ################################################################################
90 print "Building list of database files...";
91 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id")
96 filename = os.path.abspath(i[0] + i[1]);
97 db_files[filename] = "";
98 if os.access(filename, os.R_OK) == 0:
99 utils.warn("'%s' doesn't exist." % (filename));
101 filename = Cnf["Dir::Override"]+'override.unreferenced';
102 if os.path.exists(filename):
103 file = utils.open_file(filename);
104 for filename in file.readlines():
105 filename = filename[:-1];
106 excluded[filename] = "";
108 print "Checking against existent files...";
110 os.path.walk(Cnf["Dir::Root"]+'dists/', process_dir, None);
113 print "%s wasted..." % (utils.size_type(waste));
115 ################################################################################
120 for component in Cnf.SubTree("Component").List():
121 if component == "mixed":
123 component = component.lower();
124 list_filename = '%s%s_%s_source.list' % (Cnf["Dir::Lists"], suite, component);
125 list_file = utils.open_file(list_filename);
126 for line in list_file.readlines():
129 utils.parse_changes(file, dsc_whitespace_rules=1);
130 except utils.invalid_dsc_format_exc, line:
131 utils.warn("syntax error in .dsc file '%s', line %s." % (file, line));
135 utils.warn("Found %s invalid .dsc files." % (count));
137 ################################################################################
139 def check_override():
140 for suite in [ "stable", "unstable" ]:
142 print "-------------"
144 suite_id = db_access.get_suite_id(suite);
145 q = projectB.query("""
146 SELECT DISTINCT b.package FROM binaries b, bin_associations ba
147 WHERE b.id = ba.bin AND ba.suite = %s AND NOT EXISTS
148 (SELECT 1 FROM override o WHERE o.suite = %s AND o.package = b.package)"""
149 % (suite_id, suite_id));
151 q = projectB.query("""
152 SELECT DISTINCT s.source FROM source s, src_associations sa
153 WHERE s.id = sa.source AND sa.suite = %s AND NOT EXISTS
154 (SELECT 1 FROM override o WHERE o.suite = %s and o.package = s.source)"""
155 % (suite_id, suite_id));
158 ################################################################################
160 # Ensure that the source files for any given package is all in one
161 # directory so that 'apt-get source' works...
163 def check_source_in_one_dir():
164 # Not the most enterprising method, but hey...
166 q = projectB.query("SELECT id FROM source;");
167 for i in q.getresult():
169 q2 = projectB.query("""
170 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"""
175 for j in q2.getresult():
176 filename = j[0] + j[1];
177 path = os.path.dirname(filename);
180 first_filename = filename;
181 elif first_path != path:
182 symlink = path + '/' + os.path.basename(first_filename);
183 if not os.path.exists(symlink):
185 print "WOAH, we got a live one here... %s [%s] {%s}" % (filename, source_id, symlink);
188 print "Found %d source packages where the source is not all in one directory." % (broken_count);
190 ################################################################################
193 print "Getting file information from database...";
194 q = projectB.query("SELECT l.path, f.filename, f.md5sum, f.size FROM files f, location l WHERE f.location = l.id")
197 print "Checking file md5sums & sizes...";
199 filename = os.path.abspath(i[0] + i[1]);
203 file = utils.open_file(filename);
205 utils.warn("can't open '%s'." % (filename));
207 md5sum = apt_pkg.md5sum(file);
208 size = os.stat(filename)[stat.ST_SIZE];
209 if md5sum != db_md5sum:
210 utils.warn("**WARNING** md5sum mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, md5sum, db_md5sum));
212 utils.warn("**WARNING** size mismatch for '%s' ('%s' [current] vs. '%s' [db])." % (filename, size, db_size));
216 ################################################################################
218 # Check all files for timestamps in the future; common from hardware
219 # (e.g. alpha) which have far-future dates as their default dates.
221 def Ent(Kind,Name,Link,Mode,UID,GID,Size,MTime,Major,Minor):
224 if MTime > current_time:
225 future_files[current_file] = MTime;
226 print "%s: %s '%s','%s',%u,%u,%u,%u,%u,%u,%u" % (current_file, Kind,Name,Link,Mode,UID,GID,Size, MTime, Major, Minor);
228 def check_timestamps():
231 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id AND f.filename ~ '.deb$'")
236 filename = os.path.abspath(i[0] + i[1]);
237 if os.access(filename, os.R_OK):
238 file = utils.open_file(filename);
239 current_file = filename;
240 sys.stderr.write("Processing %s.\n" % (filename));
241 apt_inst.debExtract(file,Ent,"control.tar.gz");
243 apt_inst.debExtract(file,Ent,"data.tar.gz");
245 print "Checked %d files (out of %d)." % (count, len(db_files.keys()));
247 ################################################################################
249 def check_missing_tar_gz_in_dsc():
252 print "Building list of database files...";
253 q = projectB.query("SELECT l.path, f.filename FROM files f, location l WHERE f.location = l.id AND f.filename ~ '.dsc$'");
256 print "Checking %d files..." % len(ql);
258 print "No files to check."
260 filename = os.path.abspath(i[0] + i[1]);
262 # NB: don't enforce .dsc syntax
263 dsc = utils.parse_changes(filename);
265 utils.fubar("error parsing .dsc file '%s'." % (filename));
266 dsc_files = utils.build_file_list(dsc, is_a_dsc=1);
268 for file in dsc_files.keys():
269 m = utils.re_issource.match(file);
271 utils.fubar("%s not recognised as source." % (file));
273 if type == "orig.tar.gz" or type == "tar.gz":
276 utils.warn("%s has no .tar.gz in the .dsc file." % (file));
280 utils.warn("Found %s invalid .dsc files." % (count));
283 ################################################################################
285 def validate_sources(suite, component):
286 filename = "%s/dists/%s/%s/source/Sources.gz" % (Cnf["Dir::Root"], suite, component);
287 print "Processing %s..." % (filename);
288 # apt_pkg.ParseTagFile needs a real file handle and can't handle a GzipFile instance...
289 temp_filename = utils.temp_filename();
290 (result, output) = commands.getstatusoutput("gunzip -c %s > %s" % (filename, temp_filename));
292 sys.stderr.write("Gunzip invocation failed!\n%s\n" % (output));
294 sources = utils.open_file(temp_filename);
295 Sources = apt_pkg.ParseTagFile(sources);
296 while Sources.Step():
297 source = Sources.Section.Find('Package');
298 directory = Sources.Section.Find('Directory');
299 files = Sources.Section.Find('Files');
300 for i in files.split('\n'):
301 (md5, size, name) = i.split();
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);
319 os.unlink(temp_filename);
321 ########################################
323 def validate_packages(suite, component, architecture):
324 filename = "%s/dists/%s/%s/binary-%s/Packages" \
325 % (Cnf["Dir::Root"], suite, component, architecture);
326 print "Processing %s..." % (filename);
327 packages = utils.open_file(filename);
328 Packages = apt_pkg.ParseTagFile(packages);
329 while Packages.Step():
330 filename = "%s/%s" % (Cnf["Dir::Root"], Packages.Section.Find('Filename'));
331 if not os.path.exists(filename):
332 print "W: %s missing." % (filename);
335 ########################################
337 def check_indices_files_exist():
338 for suite in [ "stable", "testing", "unstable" ]:
339 for component in Cnf.ValueList("Suite::%s::Components" % (suite)):
340 architectures = Cnf.ValueList("Suite::%s::Architectures" % (suite));
341 for arch in map(string.lower, architectures):
343 validate_sources(suite, component);
347 validate_packages(suite, component, arch);
349 ################################################################################
351 def check_files_not_symlinks():
352 print "Building list of database files... ",;
353 before = time.time();
354 q = projectB.query("SELECT l.path, f.filename, f.id FROM files f, location l WHERE f.location = l.id")
355 print "done. (%d seconds)" % (int(time.time()-before));
356 q_files = q.getresult();
359 # q = projectB.query("SELECT l.path, c.name, l.id FROM location l, component c WHERE l.component = c.id");
360 # for i in q.getresult():
361 # path = os.path.normpath(i[0] + i[1]);
362 # locations[path] = (i[0], i[2]);
364 # q = projectB.query("BEGIN WORK");
366 filename = os.path.normpath(i[0] + i[1]);
368 if os.access(filename, os.R_OK) == 0:
369 utils.warn("%s: doesn't exist." % (filename));
371 if os.path.islink(filename):
372 utils.warn("%s: is a symlink." % (filename));
373 # You probably don't want to use the rest of this...
374 # print "%s: is a symlink." % (filename);
375 # dest = os.readlink(filename);
376 # if not os.path.isabs(dest):
377 # dest = os.path.normpath(os.path.join(os.path.dirname(filename), dest));
378 # print "--> %s" % (dest);
379 # # Determine suitable location ID
380 # # [in what must be the suckiest way possible?]
381 # location_id = None;
382 # for path in locations.keys():
383 # if dest.find(path) == 0:
384 # (location, location_id) = locations[path];
386 # if not location_id:
387 # utils.fubar("Can't find location for %s (%s)." % (dest, filename));
388 # new_filename = dest.replace(location, "");
389 # q = projectB.query("UPDATE files SET filename = '%s', location = %s WHERE id = %s" % (new_filename, location_id, file_id));
390 # q = projectB.query("COMMIT WORK");
392 ################################################################################
394 def chk_bd_process_dir (unused, dirname, filenames):
395 for name in filenames:
396 if not name.endswith(".dsc"):
398 filename = os.path.abspath(dirname+'/'+name);
399 dsc = utils.parse_changes(filename);
400 for field_name in [ "build-depends", "build-depends-indep" ]:
401 field = dsc.get(field_name);
404 apt_pkg.ParseSrcDepends(field);
406 print "E: [%s] %s: %s" % (filename, field_name, field);
409 ################################################################################
411 def check_build_depends():
412 os.path.walk(Cnf["Dir::Root"], chk_bd_process_dir, None);
414 ################################################################################
417 global Cnf, projectB, db_files, waste, excluded;
419 Cnf = utils.get_conf();
420 Arguments = [('h',"help","Tea::Options::Help")];
422 if not Cnf.has_key("Tea::Options::%s" % (i)):
423 Cnf["Tea::Options::%s" % (i)] = "";
425 args = apt_pkg.ParseCommandLine(Cnf, Arguments, sys.argv);
427 Options = Cnf.SubTree("Tea::Options")
432 utils.warn("tea requires at least one argument");
435 utils.warn("tea accepts only one argument");
437 mode = args[0].lower();
439 projectB = pg.connect(Cnf["DB::Name"], Cnf["DB::Host"], int(Cnf["DB::Port"]));
440 db_access.init(Cnf, projectB);
442 if mode == "md5sums":
444 elif mode == "files":
446 elif mode == "dsc-syntax":
448 elif mode == "missing-overrides":
450 elif mode == "source-in-one-dir":
451 check_source_in_one_dir();
452 elif mode == "timestamps":
454 elif mode == "tar-gz-in-dsc":
455 check_missing_tar_gz_in_dsc();
456 elif mode == "validate-indices":
457 check_indices_files_exist();
458 elif mode == "files-not-symlinks":
459 check_files_not_symlinks();
460 elif mode == "validate-builddeps":
461 check_build_depends();
463 utils.warn("unknown mode '%s'" % (mode));
466 ################################################################################
468 if __name__ == '__main__':