]> git.decadent.org.uk Git - dak.git/blobdiff - dak/contents.py
Revert "for the time we test, modify the path to a non-world-visible one"
[dak.git] / dak / contents.py
index 519ebba846bbb50e734f36ae2fd1556589033dec..3444071cdb34a9860f5f62f5ae7c0427e3830332 100644 (file)
@@ -42,6 +42,7 @@ import math
 import gzip
 import apt_pkg
 from daklib import utils
+from daklib.binary import Binary
 from daklib.config import Config
 from daklib.dbconn import DBConn
 ################################################################################
@@ -70,10 +71,10 @@ OPTIONS
         supress all output but errors
 
      -s, --suite={stable,testing,unstable,...}
-        only operate on a signle suite
+        only operate on a single suite
 
      -a, --arch={i386,amd64}
-        only operate on a signle architecture
+        only operate on a single architecture
 """
     sys.exit(exit_code)
 
@@ -82,20 +83,21 @@ OPTIONS
 # where in dak.conf all of our configuration will be stowed
 
 options_prefix = "Contents"
-options_prefix = "%s::Opitons" % options_prefix
+options_prefix = "%s::Options" % options_prefix
 
 log = logging.getLogger()
 
 ################################################################################
 
-latin1_q = """SET CLIENT_ENCODING TO 'LATIN1'"""
-
+# get all the arches delivered for a given suite
+# this should probably exist somehere common
 arches_q = """PREPARE arches_q as
               SELECT s.architecture, a.arch_string
               FROM suite_architectures s
               JOIN architecture a ON (s.architecture=a.id)
                   WHERE suite = $1"""
 
+# find me the .deb for a given binary id
 debs_q = """PREPARE debs_q as
               SELECT b.id, f.filename FROM bin_assoc_by_arch baa
               JOIN binaries b ON baa.bin=b.id
@@ -103,11 +105,13 @@ debs_q = """PREPARE debs_q as
               WHERE suite = $1
                   AND arch = $2"""
 
+# ask if we already have contents associated with this binary
 olddeb_q = """PREPARE olddeb_q as
               SELECT 1 FROM content_associations
               WHERE binary_pkg = $1
               LIMIT 1"""
 
+# find me all of the contents for a given .deb
 contents_q = """PREPARE contents_q as
               SELECT (p.path||'/'||n.file) AS fn,
                       comma_separated_list(s.section||'/'||b.package)
@@ -126,6 +130,7 @@ contents_q = """PREPARE contents_q as
               GROUP BY fn
               ORDER BY fn"""
 
+# find me all of the contents for a given .udeb
 udeb_contents_q = """PREPARE udeb_contents_q as
               SELECT (p.path||'/'||n.file) as fn,
                       comma_separated_list(s.section||'/'||b.package)
@@ -144,6 +149,25 @@ udeb_contents_q = """PREPARE udeb_contents_q as
               GROUP BY fn
               ORDER BY fn"""
 
+# clear out all of the temporarily stored content associations
+# this should be run only after p-a has run.  after a p-a
+# run we should have either accepted or rejected every package
+# so there should no longer be anything in the queue
+remove_pending_contents_cruft_q = """DELETE FROM pending_content_associations"""
+
+# delete any filenames we are storing which have no binary associated with them
+remove_filename_cruft_q = """DELETE FROM content_file_names
+                             WHERE id IN (SELECT cfn.id FROM content_file_names cfn
+                                          LEFT JOIN content_associations ca
+                                            ON ca.filename=cfn.id
+                                          WHERE ca.id IS NULL)"""
+
+# delete any paths we are storing which have no binary associated with them
+remove_filepath_cruft_q = """DELETE FROM content_file_paths
+                             WHERE id IN (SELECT cfn.id FROM content_file_paths cfn
+                                          LEFT JOIN content_associations ca
+                                             ON ca.filepath=cfn.id
+                                          WHERE ca.id IS NULL)"""
 class Contents(object):
     """
     Class capable of generating Contents-$arch.gz files
@@ -155,21 +179,24 @@ class Contents(object):
         self.header = None
 
     def _getHeader(self):
-        # Internal method to return the header for Contents.gz files
+        """
+        Internal method to return the header for Contents.gz files
+
+        This is boilerplate which explains the contents of the file and how
+        it can be used.
+        """
         if self.header == None:
             if Config().has_key("Contents::Header"):
                 try:
                     h = open(os.path.join( Config()["Dir::Templates"],
                                            Config()["Contents::Header"] ), "r")
                     self.header = h.read()
-                    print( "header: %s" % self.header )
                     h.close()
                 except:
-                    log.error( "error openeing header file: %d\n%s" % (Config()["Contents::Header"],
-                                                                       traceback.format_exc() ))
+                    log.error( "error opening header file: %d\n%s" % (Config()["Contents::Header"],
+                                                                      traceback.format_exc() ))
                     self.header = False
             else:
-                print( "no header" )
                 self.header = False
 
         return self.header
@@ -178,8 +205,15 @@ class Contents(object):
     _goal_column = 54
 
     def _write_content_file(self, cursor, filename):
-        # Internal method for writing all the results to a given file
-        f = gzip.open(Config()["Dir::Root"] + filename, "w")
+        """
+        Internal method for writing all the results to a given file.
+        The cursor should have a result set generated from a query already.
+        """
+        filepath = Config()["Contents::Root"] + filename
+        filedir = os.path.dirname(filepath)
+        if not os.path.isdir(filedir):
+            os.makedirs(filedir)
+        f = gzip.open(filepath, "w")
         try:
             header = self._getHeader()
 
@@ -192,7 +226,7 @@ class Contents(object):
                     return
 
                 num_tabs = max(1,
-                               int( math.ceil( (self._goal_column - len(contents[0])) / 8) ) )
+                               int(math.ceil((self._goal_column - len(contents[0])) / 8)))
                 f.write(contents[0] + ( '\t' * num_tabs ) + contents[-1] + "\n")
 
         finally:
@@ -200,20 +234,14 @@ class Contents(object):
 
     def cruft(self):
         """
-        remove files/paths from the DB which are no longer referenced by binaries
+        remove files/paths from the DB which are no longer referenced
+        by binaries and clean the temporary table
         """
         cursor = DBConn().cursor();
         cursor.execute( "BEGIN WORK" )
-        cursor.execute( """DELETE FROM content_file_names
-                           WHERE id IN (SELECT cfn.id FROM content_file_names cfn
-                                        LEFT JOIN content_associations ca
-                                            ON ca.filename=cfn.id
-                                        WHERE ca.id IS NULL)""" );
-        cursor.execute( """DELETE FROM content_file_paths
-                           WHERE id IN (SELECT cfn.id FROM content_file_paths cfn
-                                        LEFT JOIN content_associations ca
-                                            ON ca.filepath=cfn.id
-                                        WHERE ca.id IS NULL)""" );
+        cursor.execute( remove_pending_contents_cruft_q )
+        cursor.execute( remove_filename_cruft_q )
+        cursor.execute( remove_filepath_cruft_q )
         cursor.execute( "COMMIT" )
 
 
@@ -224,7 +252,6 @@ class Contents(object):
         pooldir = Config()[ 'Dir::Pool' ]
 
         cursor = DBConn().cursor();
-        cursor.execute( latin1_q )
         cursor.execute( debs_q )
         cursor.execute( olddeb_q )
         cursor.execute( arches_q )
@@ -238,26 +265,27 @@ class Contents(object):
             for arch_id in arch_list:
                 cursor.execute( "EXECUTE debs_q(%d, %d)" % ( suite_id, arch_id[0] ) )
 
-                debs = cursor.fetchall()
                 count = 0
-                for deb in debs:
+                while True:
+                    deb = cursor.fetchone()
+                    if not deb:
+                        break
                     count += 1
-                    cursor.execute( "EXECUTE olddeb_q(%d)" % (deb[0] ) )
-                    old = cursor.fetchone()
+                    cursor1 = DBConn().cursor();
+                    cursor1.execute( "EXECUTE olddeb_q(%d)" % (deb[0] ) )
+                    old = cursor1.fetchone()
                     if old:
                         log.debug( "already imported: %s" % deb[1] )
                     else:
                         debfile = os.path.join( pooldir, deb[1] )
                         if os.path.exists( debfile ):
-                            contents = utils.generate_contents_information( debfile )
-                            DBConn().insert_content_paths(deb[0], contents)
-                            log.info( "imported (%d/%d): %s" % (count,len(debs),deb[1] ) )
+                            Binary(debfile).scan_package( deb[0] )
                         else:
                             log.error( "missing .deb: %s" % deb[1] )
 
     def generate(self):
         """
-        Generate Contents-$arch.gz files for every aviailable arch in each given suite.
+        Generate Contents-$arch.gz files for every available arch in each given suite.
         """
         cursor = DBConn().cursor();
 
@@ -280,24 +308,21 @@ class Contents(object):
 
             # The MORE fun part. Ok, udebs need their own contents files, udeb, and udeb-nf (not-free)
             # This is HORRIBLY debian specific :-/
-            # First off, udeb
-            section_id = DBConn().get_section_id('debian-installer') # all udebs should be here)
-            if section_id != -1:
-                cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d)" % (section_id, suite_id, suite_id))
-                self._write_content_file(cursor, "dists/%s/Contents-udeb.gz" % suite)
+            for section_id, fn_pattern in [("debian-installer","dists/%s/Contents-udeb.gz"),
+                                           ("non-free/debian-installer", "dists/%s/Contents-udeb-nf.gz")]:
 
-            # Once more, with non-free
-            section_id = DBConn().get_section_id('non-free/debian-installer') # all udebs should be here)
-
-            if section_id != -1:
-                cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d)" % (section_id, suite_id, suite_id))
-                self._write_content_file(cursor, "dists/%s/Contents-udeb-nf.gz" % suite)
+                section_id = DBConn().get_section_id(section_id) # all udebs should be here)
+                if section_id != -1:
+                    cursor.execute("EXECUTE udeb_contents_q(%d,%d,%d)" % (section_id, suite_id, suite_id))
+                    self._write_content_file(cursor, fn_pattern % suite)
 
 
 ################################################################################
 
     def _suites(self):
-        # return a list of suites to operate on
+        """
+        return a list of suites to operate on
+        """
         if Config().has_key( "%s::%s" %(options_prefix,"Suite")):
             suites = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Suite")])
         else:
@@ -306,7 +331,9 @@ class Contents(object):
         return suites
 
     def _arches(self, cursor, suite):
-        # return a list of archs to operate on
+        """
+        return a list of archs to operate on
+        """
         arch_list = [ ]
         if Config().has_key( "%s::%s" %(options_prefix,"Arch")):
             archs = utils.split_args(Config()[ "%s::%s" %(options_prefix,"Arch")])
@@ -349,7 +376,7 @@ def main():
         level=logging.DEBUG
 
 
-    logging.basicConfig( level=logging.DEBUG,
+    logging.basicConfig( level=level,
                          format='%(asctime)s %(levelname)s %(message)s',
                          stream = sys.stderr )