Initial commit

First commit of gitdm to the new repo.  Call it version 0.10 or something
silly like that.

9 files changed, 1191 insertions, 0 deletions

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..f3d74a9
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+*.pyc
+*~
diff --git a/COPYING b/COPYING
new file mode 100644
index 0000000..fe3eb43
--- /dev/null
+++ b/COPYING
@@ -0,0 +1,2 @@
+The code in this directory can be distributed under the terms of the GNU
+General Public License, version 2.  
diff --git a/ConfigFile.py b/ConfigFile.py
new file mode 100644
index 0000000..39310fb
--- /dev/null
+++ b/ConfigFile.py
@@ -0,0 +1,110 @@
+#
+# Stuff for dealing with configuration files.
+#
+import sys, re, datetime
+import database
+
+#
+# Read a line and strip out junk.
+#
+def ReadConfigLine (file):
+    line = file.readline ()
+    if not line:
+        return None
+    line = line.split('#')[0] # Get rid of any comments
+    line = line.strip () # and extra white space
+    if len (line) == 0: # we got rid of everything
+        return ReadConfigLine (file)
+    return line
+
+#
+# Give up and die.
+#
+def croak (message):
+    sys.stderr.write (message + '\n')
+    sys.exit (1)
+
+#
+# Read a list of email aliases.
+#
+def ReadEmailAliases (name):
+    try:
+        file = open (name, 'r')
+    except IOError:
+        croak ('Unable to open email alias file %s' % (name))
+    line = ReadConfigLine (file)
+    while line:
+        sline = line.split ()
+        if len (sline) != 2:
+            croak ('Funky email alias line "%s"' % (line))
+        if sline[0].index ('@') <= 0 or sline[1].index ('@') <= 0:
+            croak ('Non-addresses in email alias "%s"' % (line))
+        database.AddEmailAlias (sline[0], sline[1])
+        line = ReadConfigLine (file)
+    file.close ()
+
+#
+# The Email/Employer map
+#
+EMMpat = re.compile (r'^([^\s]+)\s+([^<]+)\s*(<\s*(\d+-\d+-\d+)\s*)?$')
+
+def ReadEmailEmployers (name):
+    try:
+        file = open (name, 'r')
+    except IOError:
+        croak ('Unable to open email/employer file %s' % (name))
+    line = ReadConfigLine (file)
+    while line:
+        m = EMMpat.match (line)
+        if not m:
+            croak ('Funky email/employer line "%s"' % (line))
+        email = m.group (1)
+        company = m.group (2).strip ()
+        enddate = ParseDate (m.group (4))
+        database.AddEmailEmployerMapping (email, company, enddate)
+        line = ReadConfigLine (file)
+    file.close ()
+
+def ParseDate (cdate):
+    if not cdate:
+        return None
+    sdate = cdate.split ('-')
+    return datetime.date (int (sdate[0]), int (sdate[1]), int (sdate[2]))
+
+
+def ReadGroupMap (fname, employer):
+    try:
+        file = open (fname, 'r')
+    except IOError:
+        croak ('Unable to open group map file %s' % (fname))
+    line = ReadConfigLine (file)
+    while line:
+        database.AddEmailEmployerMapping (line, employer)
+        line = ReadConfigLine (file)
+    file.close ()
+
+#
+# Read an overall config file.
+#
+def ConfigFile (name):
+    try:
+        file = open (name, 'r')
+    except IOError:
+        croak ('Unable to open config file %s' % (name))
+    line = ReadConfigLine (file)
+    while line:
+        sline = line.split ()
+        if len (sline) < 2:
+            croak ('Funky config line: "%s"' % (line))
+        if sline[0] == 'EmailAliases':
+            ReadEmailAliases (sline[1])
+        elif sline[0] == 'EmailMap':
+            ReadEmailEmployers (sline[1])
+        elif sline[0] == 'GroupMap':
+            if len (sline) != 3:
+                croak ('Funky group map line "%s"' % (line))
+            ReadGroupMap (sline[1], sline[2])
+        else:
+            croak ('Unrecognized config line: "%s"' % (line))
+        line = ReadConfigLine (file)
+        
diff --git a/README b/README
new file mode 100644
index 0000000..62c8d31
--- /dev/null
+++ b/README
@@ -0,0 +1,107 @@
+The code in this directory makes up the "git data miner," a simple hack
+which attempts to figure things out from the revision history in a git
+repository. 
+
+RUNNING GITDM
+
+Run it like this:
+
+   git log -p -M [details] | gitdm [options]
+
+The [details] tell git which changesets are of interest; the [options] can
+be:
+
+	-a	If a patch contains signoff lines from both Andrew Morton 
+		and Linus Torvalds, omit Linus's.
+
+	-c file Specify the name of the gitdm configuration file.  
+	   	By default, "./gitdm.config" is used.
+
+	-d	Omit the developer reports, giving employer information
+         	only. 
+
+	-D	Rather than create the usual statistics, create a 
+		file providing lines changed per day, suitable for
+		feeding to a tool like gnuplot.
+
+	-h file	Generate HTML output to the given file
+
+	-l num	Only list the top <num> entries in each report.
+
+	-o file	Write text output to the given file (default is stdout).
+
+	-r pat	Only generate statistics for changes to files whose 
+	   	name matches the given regular expression.
+
+	-s	Ignore Signed-off-by lines which match the author of 
+		each patch.
+
+	-u 	Group all unknown developers under the "(Unknown)"
+ 	        employer. 
+
+	-z 	Dump out the hacker database to "database.dump".
+
+A typical command line used to generate the "who write 2.6.x" LWN articles
+looks like:
+
+    git log -p -M v2.6.19..v2.6.20 | \
+	gitdm -u -s -a -o results -h results.html
+
+
+CONFIGURATION FILE
+
+The main purpose of the configuration file is to direct the mapping of
+email addresses onto employers.  Please note that the config file parser is
+exceptionally stupid and unrobust at this point, but it gets the job done.  
+
+Blank lines and lines beginning with "#" are ignored.  Everything else
+specifies a file with some sort of mapping:
+
+EmailAliases file
+
+	Developers often post code under a number of different email
+	addresses, but it can be desirable to group them all together in
+	the statistics.  An EmailAliases file just contains a bunch of
+	lines of the form:
+
+		alias@address  canonical@address
+
+	Any patches originating from alias@address will be treated as if
+	they had come from canonical@address.
+
+
+EmailMap file
+
+	Map email addresses onto employers.  These files contain lines
+	like:
+
+		[user@]domain  employer  [< yyyy-mm-dd]
+
+	If the "user@" portion is missing, all email from the given domain
+	will be treated as being associated with the given employer.  If a
+	date is provided, the entry is only valid up to that date;
+	otherwise it is considered valid into the indefinite future.  This
+	feature can be useful for properly tracking developers' work when
+	they change employers but do not change email addresses.
+
+
+GroupMap file employer
+
+	This is a variant of EmailMap provided for convenience; it contains
+	email addresses only, all of which are associated with the given
+	employer.
+
+
+NOTES AND CREDITS
+
+Gitdm was written by Jonathan Corbet; many useful contributions have come
+from Greg Kroah-Hartman.
+
+Please note that this tool is provided in the hope that it will be useful,
+but it is not put forward as an example of excellence in design or
+implementation.  Hacking on gitdm tends to stop the moment it performs
+whatever task is required of it at the moment.  Patches to make it less
+hacky, less ugly, and more robust are welcome.
+
+Jonathan Corbet
+corbet@lwn.net
diff --git a/database.py b/database.py
new file mode 100644
index 0000000..edb54cf
--- /dev/null
+++ b/database.py
@@ -0,0 +1,202 @@
+#
+# The "database".
+#
+
+#
+# This code is part of the LWN git data miner.
+#
+# Copyright 2007 LWN.net
+# Copyright 2007 Jonathan Corbet <corbet@lwn.net>
+#
+# This file may be distributed under the terms of the GNU General
+# Public License, version 2.
+import sys, datetime
+
+
+class Hacker:
+    def __init__ (self, name, id, elist, email):
+        self.name = name
+        self.id = id
+        self.employer = [ elist ]
+        self.email = [ email ]
+        self.added = self.removed = 0
+        self.patches = [ ]
+        self.signoffs = [ ]
+
+    def addemail (self, email, elist):
+        self.email.append (email)
+        self.employer.append (elist)
+        HackersByEmail[email] = self
+
+    def emailemployer (self, email, date):
+        for i in range (0, len (self.email)):
+            if self.email[i] == email:
+                for edate, empl in self.employer[i]:
+                    if edate > date:
+                        return empl
+        print 'OOPS.  ', self.name, self.employer, self.email, email, date
+        return None # Should not happen
+
+    def addpatch (self, patch):
+        self.added += patch.added
+        self.removed += patch.removed
+        self.patches.append (patch)
+
+    def addsob (self, patch):
+        self.signoffs.append (patch)
+
+HackersByName = { }
+HackersByEmail = { }
+HackersByID = { }
+MaxID = 0
+
+def StoreHacker (name, elist, email):
+    global MaxID
+
+    id = MaxID
+    MaxID += 1
+    h = Hacker (name, id, elist, email)
+    HackersByName[name] = h
+    HackersByEmail[email] = h
+    HackersByID[id] = h
+    return h
+
+def LookupEmail (addr):
+    try:
+        return HackersByEmail[addr]
+    except KeyError:
+        return None
+
+def LookupName (name):
+    try:
+        return HackersByName[name]
+    except KeyError:
+        return None
+
+def LookupID (id):
+    try:
+        return HackersByID[id]
+    except KeyError:
+        return None
+
+def AllHackers ():
+    return HackersByID.values ()
+#    return [h for h in HackersByID.values ()] #  if (h.added + h.removed) > 0]
+
+def DumpDB ():
+    out = open ('database.dump', 'w')
+    names = HackersByName.keys ()
+    names.sort ()
+    for name in names:
+        h = HackersByName[name]
+        out.write ('%4d %s %d p (+%d -%d) sob: %d\n' % (h.id, h.name,
+                                                        len (h.patches),
+                                                        h.added, h.removed,
+                                                        len (h.signoffs)))
+        for i in range (0, len (h.email)):
+            out.write ('\t%s -> \n' % (h.email[i]))
+            for date, empl in h.employer[i]:
+                out.write ('\t\t %d-%d-%d %s\n' % (date.year, date.month, date.day,
+                                                 empl.name))
+
+#
+# Employer info.
+#
+class Employer:
+    def __init__ (self, name):
+        self.name = name
+        self.added = self.removed = self.count = self.changed = 0
+        self.sobs = 0
+        self.hackers = [ ]
+
+    def AddCSet (self, patch):
+        self.added += patch.added
+        self.removed += patch.removed
+        self.changed += max(patch.added, patch.removed)
+        self.count += 1
+        if patch.author not in self.hackers:
+            self.hackers.append (patch.author)
+
+    def AddSOB (self):
+        self.sobs += 1
+
+Employers = { }
+
+def GetEmployer (name):
+    try:
+        return Employers[name]
+    except KeyError:
+        e = Employer (name)
+        Employers[name] = e
+        return e
+
+def AllEmployers ():
+    return Employers.values ()
+
+#
+# The email map.
+#
+EmailAliases = { }
+
+def AddEmailAlias (variant, canonical):
+    if EmailAliases.has_key (variant):
+        sys.stderr.write ('Duplicate email alias for %s\n' % (variant))
+    EmailAliases[variant] = canonical
+
+def RemapEmail (email):
+    email = email.lower ()
+    try:
+        return EmailAliases[email]
+    except KeyError:
+        return email
+
+#
+# Email-to-employer mapping.
+#
+EmailToEmployer = { }
+nextyear = datetime.date.today () + datetime.timedelta (days = 365)
+
+def AddEmailEmployerMapping (email, employer, end = nextyear):
+    if end is None:
+        end = nextyear
+    email = email.lower ()
+    empl = GetEmployer (employer)
+    try:
+        l = EmailToEmployer[email]
+        print email, l
+        for i in range (0, len(l)):
+            date, xempl = l[i]
+            if date == end:  # probably both nextyear
+                print 'WARNING: duplicate email/empl for %s' % (email)
+            if date > end:
+                l.insert (i, (end, empl))
+                return
+        l.append ((end, empl))
+    except KeyError:
+        EmailToEmployer[email] = [(end, empl)]
+
+def MapToEmployer (email, unknown = 0):
+    email = email.lower ()
+    try:
+        return EmailToEmployer[email]
+    except KeyError:
+        pass
+    namedom = email.split ('@')
+    if len (namedom) < 2:
+        print 'Oops...funky email %s' % email
+        return [(nextyear, GetEmployer ('Funky'))]
+    s = namedom[1].split ('.')
+    for dots in range (len (s) - 2, -1, -1):
+        addr = '.'.join (s[dots:])
+        try:
+            return EmailToEmployer[addr]
+        except KeyError:
+            pass
+    if unknown:
+        return [(nextyear, GetEmployer ('(Unknown)'))]
+    return [(nextyear, GetEmployer (email))]
+
+
+def LookupEmployer (email, mapunknown = 0):
+    elist = MapToEmployer (email, mapunknown)
+    return elist # GetEmployer (ename)
diff --git a/gitdm b/gitdm
new file mode 100755
index 0000000..32f27c8
--- /dev/null
+++ b/gitdm
@@ -0,0 +1,499 @@
+#!/usr/bin/python
+#
+
+#
+# This code is part of the LWN git data miner.
+#
+# Copyright 2007 LWN.net
+# Copyright 2007 Jonathan Corbet <corbet@lwn.net>
+#
+# This file may be distributed under the terms of the GNU General
+# Public License, version 2.
+
+
+import database, ConfigFile
+import getopt, datetime
+import os, re, sys, rfc822, string
+
+#
+# Some people, when confronted with a problem, think "I know, I'll use regular
+# expressions." Now they have two problems.
+#    -- Jamie Zawinski
+#
+Pcommit = re.compile (r'^commit ([0-9a-f]+)$')
+Pauthor = re.compile (r'^Author: ([^<]+)\s<([^>]+)>$')
+Psob = re.compile (r'Signed-off-by:\s+([^<]+)\s+<([^>]+)>')
+Pmerge = re.compile (r'^Merge:.*$')
+Padd = re.compile (r'^\+[^\+].*$')
+Prem = re.compile (r'^-[^-].*$')
+Pdate = re.compile (r'^(Commit)?Date:\s+(.*)$')
+Pfilea = re.compile (r'^---\s+(.*)$')
+Pfileb = re.compile (r'^\+\+\+\s+(.*)$')
+
+class patch:
+    pass
+
+
+#
+# Control options.
+#
+Outfile = sys.stdout
+ListCount = 999999
+MapUnknown = 0
+DevReports = 1
+DateStats = 0
+AuthorSOBs = 1
+FileFilter = None
+AkpmOverLt = 0
+DumpDB = 0
+CFName = 'gitdm.config'
+#
+# Options:
+#
+# -a		Andrew Morton's signoffs shadow Linus's
+# -c cfile	Specify a configuration file
+# -d		Output individual developer stats
+# -D		Output date statistics
+# -h hfile	HTML output to hfile
+# -l count	Maximum length for output lists
+# -o file	File for text output
+# -r pattern	Restrict to files matching pattern
+# -s		Ignore author SOB lines
+# -u		Map unknown employers to '(Unknown)'
+# -z		Dump out the hacker database at completion
+
+def ParseOpts ():
+    global Outfile, ListCount, MapUnknown, HTMLfile, DevReports
+    global DateStats, AuthorSOBs, FileFilter, AkpmOverLt, DumpDB
+    global CFName
+
+    opts, rest = getopt.getopt (sys.argv[1:], 'adDh:l:o:r:suz')
+    for opt in opts:
+        if opt[0] == '-a':
+            AkpmOverLt = 1
+        elif opt[0] == '-c':
+            CFName = opt[1]
+        elif opt[0] == '-d':
+            DevReports = 0
+        elif opt[0] == '-D':
+            DateStats = 1
+        elif opt[0] == '-h':
+            HTMLfile = open (opt[1], 'w')
+        elif opt[0] == '-l':
+            ListCount = int (opt[1])
+        elif opt[0] == '-o':
+            Outfile = open (opt[1], 'w')
+        elif opt[0] == '-r':
+            print 'Filter on "%s"' % (opt[1])
+            FileFilter = re.compile (opt[1])
+        elif opt[0] == '-s':
+            AuthorSOBs = 0
+        elif opt[0] == '-u':
+            MapUnknown = 1
+        elif opt[0] == '-z':
+            DumpDB = 1
+        
+
+
+def LookupStoreHacker (name, email):
+    email = database.RemapEmail (email)
+    h = database.LookupEmail (email)
+    if h: # already there
+        return h
+    elist = database.LookupEmployer (email, MapUnknown)
+    h = database.LookupName (name)
+    if h: # new email
+        h.addemail (email, elist)
+        return h
+    return database.StoreHacker(name, elist, email)
+
+#
+# Date tracking.
+#
+
+DateMap = { }
+
+def AddDateLines(date, lines):
+    if lines > 1000000:
+        print 'Skip big patch (%d)' % lines
+        return
+    dt = (date.year, date.month, date.day)
+    try:
+        DateMap[date] += lines
+    except KeyError:
+        DateMap[date] = lines
+
+def PrintDateStats():
+    dates = DateMap.keys ()
+    dates.sort ()
+    total = 0
+    datef = open ('datelc', 'w')
+    for date in dates:
+        total += DateMap[date]
+        datef.write ('%d/%02d/%02d %6d %7d\n' % (date[0], date[1], date[2],
+                                    DateMap[date], total))
+
+#
+# The core hack for grabbing the information about a changeset.
+#
+def grabpatch():
+    global NextLine, TotalAdded, TotalRemoved, TotalChanged
+    
+    while (1):
+        m = Pcommit.match (NextLine)
+        if m:
+            break;
+        NextLine = sys.stdin.readline ()
+        if not NextLine:
+            return
+
+    p = patch()
+    p.commit = m.group (1)
+    p.merge = p.added = p.removed = 0
+    p.author = LookupStoreHacker('Unknown hacker', 'unknown@hacker.net')
+    p.email = 'unknown@hacker.net'
+    p.sobs = [ ]
+    NextLine = sys.stdin.readline ()
+    ignore = (FileFilter is not None)
+    while NextLine:
+        Line = NextLine
+        #
+        # If this line starts a new commit, drop out.
+        #
+        m = Pcommit.match (Line)
+        if m:
+            break
+        NextLine = sys.stdin.readline ()
+        #
+        # Maybe it's an author line?
+        #
+        m = Pauthor.match (Line)
+        if m:
+            p.email = database.RemapEmail (m.group (2))
+            p.author = LookupStoreHacker(m.group (1), p.email)
+            continue
+        #
+        # Could be a signed-off-by:
+        #
+        m = Psob.search (Line)
+        if m:
+            email = database.RemapEmail (m.group (2))
+            sobber = LookupStoreHacker(m.group (1), email)
+            if sobber != p.author or AuthorSOBs:
+                p.sobs.append ((email, LookupStoreHacker(m.group (1), m.group (2))))
+            continue
+        #
+        # If this one is a merge, make note of the fact.
+        #
+        m = Pmerge.match (Line)
+        if m:
+            p.merge = 1
+            continue
+        #
+        # See if it's the date.
+        #
+        m = Pdate.match (Line)
+        if m:
+            dt = rfc822.parsedate(m.group (2))
+            p.date = datetime.date (dt[0], dt[1], dt[2])
+            continue
+        #
+        # If we have a file filter, check for file lines.
+        #
+        if FileFilter:
+            ignore = ApplyFileFilter (Line, ignore)
+        #
+        # OK, maybe it's part of the diff itself.
+        #
+        if not ignore:
+            if Padd.match (Line):
+                p.added += 1
+                continue
+            if Prem.match (Line):
+                p.removed += 1        
+    #
+    # Record some global information - but only if this patch had
+    # stuff which wasn't ignored.  This work should be done
+    # elsewhere,
+    #
+    if ((p.added + p.removed) > 0 or not FileFilter) and not p.merge:
+        TotalAdded += p.added
+        TotalRemoved += p.removed
+        TotalChanged += max (p.added, p.removed)
+        AddDateLines (p.date, max (p.added, p.removed))
+        empl = p.author.emailemployer (p.email, p.date)
+        empl.AddCSet (p)
+        if AkpmOverLt:
+            TrimLTSOBs (p)
+        for sobemail, sobber in p.sobs:
+            empl = sobber.emailemployer (sobemail, p.date)
+            empl.AddSOB()
+    return p
+
+
+def ApplyFileFilter (line, ignore):
+    #
+    # If this is the first file line (--- a/), set ignore one way
+    # or the other.
+    #
+    m = Pfilea.match (line)
+    if m:
+        file = m.group (1)
+        if FileFilter.search (file):
+            return 0
+        return 1
+    #
+    # For the second line, we can turn ignore off, but not on
+    #
+    m = Pfileb.match (line)
+    if m:
+        file = m.group (1)
+        if FileFilter.search (file):
+            return 0
+    return ignore
+
+#
+# If this patch is signed off by both Andrew Morton and Linus Torvalds,
+# remove the (redundant) Linus signoff.
+#
+def TrimLTSOBs (p):
+    if Linus in p.sobs and Akpm in p.sobs:
+        p.sobs.remove (Linus)
+
+#
+# HTML output support stuff.
+#
+HTMLfile = None
+HTMLclass = 0
+HClasses = ['Even', 'Odd']
+
+THead = '''<p>
+<table cellspacing=3>
+<tr><th colspan=3>%s</th></tr>
+'''
+
+
+def BeginReport (title):
+    global HTMLclass
+    
+    Outfile.write ('\n%s\n' % title)
+    if HTMLfile:
+        HTMLfile.write (THead % title)
+        HTMLclass = 0
+
+TRow = '''    <tr class="%s">
+<td>%s</td><td align="right">%d</td><td align="right">%.1f%%</td></tr>
+'''
+
+def ReportLine (text, count, pct):
+    global HTMLclass
+    if count == 0:
+        return
+    Outfile.write ('%-25s %4d (%.1f%%)\n' % (text, count, pct))
+    if HTMLfile:
+        HTMLfile.write (TRow % (HClasses[HTMLclass], text, count, pct))
+        HTMLclass ^= 1
+
+def EndReport ():
+    if HTMLfile:
+        HTMLfile.write ('</table>\n\n')
+        
+#
+# Comparison and report generation functions.
+#
+def ComparePCount (h1, h2):
+    return len (h2.patches) - len (h1.patches)
+
+def ReportByPCount (hlist):
+    hlist.sort (ComparePCount)
+    count = 0
+    BeginReport ('Developers with the most changesets')
+    for h in hlist:
+        pcount = len (h.patches)
+        changed = max(h.added, h.removed)
+        delta = h.added - h.removed
+        if pcount > 0:
+            ReportLine (h.name, pcount, (pcount*100.0)/CSCount)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+            
+def CompareLChanged (h1, h2):
+    return max(h2.added, h2.removed) - max(h1.added, h1.removed)
+
+def ReportByLChanged (hlist):
+    hlist.sort (CompareLChanged)
+    count = 0
+    BeginReport ('Developers with the most changed lines')
+    for h in hlist:
+        pcount = len (h.patches)
+        changed = max(h.added, h.removed)
+        delta = h.added - h.removed
+        if (h.added + h.removed) > 0:
+            ReportLine (h.name, changed, (changed*100.0)/TotalChanged)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+            
+def CompareLRemoved (h1, h2):
+    return (h2.removed - h2.added) - (h1.removed - h1.added)
+
+def ReportByLRemoved (hlist):
+    hlist.sort (CompareLRemoved)
+    count = 0
+    BeginReport ('Developers with the most lines removed')
+    for h in hlist:
+        pcount = len (h.patches)
+        changed = max(h.added, h.removed)
+        delta = h.added - h.removed
+        if delta < 0:
+            ReportLine (h.name, -delta, (-delta*100.0)/TotalRemoved)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+
+def CompareEPCount (e1, e2):
+    return e2.count - e1.count
+
+def ReportByPCEmpl (elist):
+    elist.sort (CompareEPCount)
+    count = 0
+    BeginReport ('Top changeset contributors by employer')
+    for e in elist:
+        if e.count != 0:
+            ReportLine (e.name, e.count, (e.count*100.0)/CSCount)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+
+
+
+def CompareELChanged (e1, e2):
+    return e2.changed - e1.changed
+
+def ReportByELChanged (elist):
+    elist.sort (CompareELChanged)
+    count = 0
+    BeginReport ('Top lines changed by employer')
+    for e in elist:
+        if e.changed != 0:
+            ReportLine (e.name, e.changed, (e.changed*100.0)/TotalChanged)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+
+
+
+def CompareSOBs (h1, h2):
+    return len (h2.signoffs) - len (h1.signoffs)
+
+def ReportBySOBs (hlist):
+    hlist.sort (CompareSOBs)
+    totalsobs = 0
+    for h in hlist:
+        totalsobs += len (h.signoffs)
+    count = 0
+    BeginReport ('Developers with the most signoffs (total %d)' % totalsobs)
+    for h in hlist:
+        scount = len (h.signoffs)
+        if scount > 0:
+            ReportLine (h.name, scount, (scount*100.0)/totalsobs)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+
+def CompareESOBs (e1, e2):
+    return e2.sobs - e1.sobs
+
+def ReportByESOBs (elist):
+    elist.sort (CompareESOBs)
+    totalsobs = 0
+    for e in elist:
+        totalsobs += e.sobs
+    count = 0
+    BeginReport ('Employers with the most signoffs (total %d)' % totalsobs)
+    for e in elist:
+        if e.sobs > 0:
+            ReportLine (e.name, e.sobs, (e.sobs*100.0)/totalsobs)
+        count += 1
+        if count >= ListCount:
+            break
+    EndReport ()
+
+#
+# Here starts the real program.  Read the config files.
+#
+ConfigFile.ConfigFile (CFName)
+
+#
+# Let's pre-seed the database with a couple of hackers
+# we want to remember.
+#
+Linus = ('torvalds@linux-foundation.org',
+         LookupStoreHacker ('Linus Torvalds', 'torvalds@linux-foundation.org'))
+Akpm = ('akpm@linux-foundation.org',
+        LookupStoreHacker ('Andrew Morton', 'akpm@linux-foundation.org'))
+
+NextLine = sys.stdin.readline ()
+TotalChanged = TotalAdded = TotalRemoved = 0
+ParseOpts ()
+
+#
+# Snarf changesets.
+#
+print 'Grabbing changesets...\r',
+sys.stdout.flush ()
+
+printcount = CSCount = 0
+while (1):
+    if (printcount % 50) == 0:
+        print 'Grabbing changesets...%d\r' % printcount,
+        sys.stdout.flush ()
+    printcount += 1
+    p = grabpatch()
+    if not p:
+        break
+    if p.added > 100000 or p.removed > 100000:
+        print 'Skipping massive add'
+        continue
+    if FileFilter and p.added == 0 and p.removed == 0:
+        continue
+    if not p.merge:
+        p.author.addpatch (p)
+        for sobemail, sob in p.sobs:
+            sob.addsob (p)
+    CSCount += 1
+print 'Grabbing changesets...done'
+
+if DumpDB:
+    database.DumpDB ()
+#
+# Say something
+#
+hlist = database.AllHackers ()
+elist = database.AllEmployers ()
+Outfile.write ('Processed %d csets from %d developers\n' % (CSCount,
+                                                            len (hlist)))
+Outfile.write ('%d employers found\n' % len (elist))
+Outfile.write ('A total of %d lines added, %d removed (delta %d)\n' %
+               (TotalAdded, TotalRemoved, TotalAdded - TotalRemoved))
+if TotalChanged == 0:
+    TotalChanged = 1 # HACK to avoid div by zero
+if DateStats:
+    PrintDateStats ()
+    sys.exit(0)
+
+if DevReports:
+    ReportByPCount (hlist)
+    ReportByLChanged (hlist)
+    ReportByLRemoved (hlist)
+    ReportBySOBs (hlist)
+ReportByPCEmpl (elist)
+ReportByELChanged (elist)
+ReportByESOBs (elist)
diff --git a/gitdm.config b/gitdm.config
new file mode 100644
index 0000000..588d6ef
--- /dev/null
+++ b/gitdm.config
@@ -0,0 +1,22 @@
+#
+# This is a sample gitdm configuration file.
+#
+
+#
+# EmailAliases lets us cope with developers who use more
+# than one address.
+#
+EmailAliases sample-config/aliases
+
+#
+# EmailMap does the main work of mapping addresses onto 
+# employers.
+#
+EmailMap sample-config/domain-map
+
+#
+# Use GroupMap to map a file full of addresses to the 
+# same employer
+#
+# GroupMap sample-config/illuminati  The Illuminati
+#
diff --git a/sample-config/aliases b/sample-config/aliases
new file mode 100644
index 0000000..8cd50db
--- /dev/null
+++ b/sample-config/aliases
@@ -0,0 +1,5 @@
+#
+# This is the email aliases file, mapping secondary addresses
+# onto a single, canonical address.
+#
+corbet@eklektix.com	corbet@lwn.net
diff --git a/sample-config/domain-map b/sample-config/domain-map
new file mode 100644
index 0000000..bbd81f7
--- /dev/null
+++ b/sample-config/domain-map
@@ -0,0 +1,242 @@
+#
+# Here is a set of mappings of domain names onto employer names.
+#
+8d.com		8D Technologies
+aconex.com	Aconex
+adaptec.com	Adaptec
+aist.go.jp	National Institute of Advanced Industrial Science and Technology
+akamai.com	Akamai Technologies
+am.sony.com	Sony
+amd.com		AMD
+analog.com	Analog Devices
+arastra.com	Arastra Inc
+arm.com		ARM
+artecdesign.ee	Artec Design
+arvoo.nl	ARVOO Engineering
+atmel.com	Atmel
+atomide.com	Atomide
+avtrex.com	Avtrex
+axis.com	Axis Communications
+azingo.com	Azingo
+balabit.com	BalaBit
+balabit.hu	BalaBit
+baslerweb.com	Basler Vision Technologies
+bluehost.com	Bluehost
+bluewatersys.com	Bluewater Systems
+broadcom.com	Broadcom
+brontes3d.com	Brontes Technologies
+bull.net	Bull SAS
+cam.ac.uk	University of Cambridge
+ccur.com	Concurrent Computer Corporation
+celunite.com	Azingo
+chelsio.com	Chelsio
+cisco.com	Cisco
+citi.umich.edu	Univ. of Michigan CITI
+clusterfs.com	Sun
+cn.fujitsu.com	Fujitsu
+compulab.co.il	CompuLab
+computergmbh.de	CC Computer Consultants
+comx.dk		ComX Networks
+conectiva.com.br	Mandriva
+coraid.com	Coraid
+cosmosbay.com	Cosmosbay~Vectis
+cozybit.com	cozybit
+cray.com	Cray
+csr.com		CSR
+cyberguard.com	Secure Computing
+cybernetics.com	Cybernetics
+data.slu.se	Uppsala University
+dave.eu		Dave S.r.l.
+de.bosch.com	Bosch
+dell.com	Dell
+denx.de		DENX Software Engineering
+devicescape.com	Devicescape
+digi.com	Digi International
+dti2.net	DTI2 - Desarrollo de la tecnologia de las comunicaciones
+edesix.com	Edesix Ltd
+elandigitalsystems.com	Elan Digital Systems
+embeddedalley.com	Embedded Alley Solutions
+empirix.com	Empirix
+emulex.com	Emulex
+endrelia.com	Endrelia
+ericsson.com	Ericsson
+fixstars.com	Fixstars Technologies
+free-electrons.com	Free Electrons
+freescale.com	Freescale
+fujitsu.com	Fujitsu
+gaisler.com	Gaisler Research
+gefanuc.com	GE Fanuc
+geomatys.fr	Geomatys
+google.com	Google
+gvs.co.yu	GVS
+hansenpartnership.com	Hansen Partnership
+harris.com	Harris Corporation
+hauppauge.com	Hauppauge
+hermes-softlab.com	HERMES SoftLab
+hevs.ch	HES-SO Valais Wallis
+highpoint-tech.com	HighPoint Technologies
+hitachi.co.jp	Hitachi
+hitachi.com	Hitachi
+hitachisoft.jp	Hitachi
+hp.com		HP
+hvsistemas.es	HV Sistemas
+ibm.com		IBM
+ibp.de		ipb (uk) Ltd.
+icplus.com.tw	IC Plus
+igel.co.jp	igel
+inl.fr		INL
+inria.fr	INRIA
+intel.com	Intel
+iram.es		IRAM
+jmicron.com	jmicron.com
+jp.fujitsu.com	Fujitsu
+katalix.com	Katalix Systems
+keyspan.com	InnoSys
+laptop.org	OLPC
+laurelnetworks.com	ECI Telecom
+linutronix.de	linutronix
+linux-foundation.org	Linux Foundation
+lippert-at.de	LiPPERT Embedded Computers GmbH
+lippertembedded.de	LiPPERT Embedded Computers GmbH
+llnl.gov	Lawrence Livermore National Laboratory
+lnxi.com	Linux Networx
+logitech.com	Logitech
+lsi.com		LSI Logic
+lsil.com	LSI Logic
+lwn.net		LWN.net
+macqel.be	Macq Electronique
+macqel.com	Macq Electronique
+mandriva.com	Mandriva
+mandriva.com.br	Mandriva
+marvell.com	Marvell
+mellanox.co.il	Mellanox
+melware.de	Cytronics & Melware
+microgate.com	MicroGate Systems
+mips.com	MIPS
+miraclelinux.com	Miracle Linux
+mn-solutions.de	M&N Solutions
+moreton.com.au	Secure Computing
+motorola.com	Motorola
+movial.fi	Movial
+mvista.com	MontaVista
+myri.com	Myricom
+namesys.com	NameSys
+nec.co.jp	NEC
+nec.com		NEC
+netapp.com	NetApp
+neterion.com	Neterion
+netxen.com	NetXen
+niif.hu		NIIF Institute
+nokia.com	Nokia
+nomadgs.com	Nomad Global Solutions
+nortel.com	Nortel
+novell.com	Novell
+ntt.co.jp	NTT
+ntts.co.jp	NTT
+nuovasystems.com	Nuova Systems
+nvidia.com	NVidia
+obsidianresearch.com	Obsidian Research
+octant-fr.com	Octant Informatique
+onelan.co.uk	ONELAN
+onstor.com	Onstor
+openedhand.com	OpenedHand
+opengridcomputing.com	Open Grid Computing
+openmoko.org	OpenMoko
+openvz.org	Parallels
+oracle.com	Oracle
+ornl.gov	Oak Ridge National Laboratory
+osdl.org	Linux Foundation
+ozlabs.org	IBM
+panasas.com	Panasas
+panasonic.com	Panasonic
+papercut.bz	PaperCut Software
+papercut.com	PaperCut Software
+parallels.com	Parallels
+pasemi.com	PA Semi Corporation
+pengutronix.de	Pengutronix
+pheonix.com	Phoeonix
+philosys.de	Philosys Software
+pikron.com	PiKRON s.r.o
+pmc-sierra.com	PMC-Sierra
+promise.com	Promise Technology
+qlogic.com	QLogic
+qumranet.com	Qumranet
+realtek.com.tw	Realtek
+redhat.com	Red Hat
+renesas.com	Renesas Technology
+rockwell.com	Rockwell
+rowland.harvard.edu	Rowland Institute, Harvard
+rtr.ca		Real-Time Remedies
+samsung.com	Samsung
+sanpeople.com	SANPeople
+savantav.com	Savant Systems
+secretlab.ca	Secretlab
+securecomputing.com	Secure Computing
+semihalf.com	Semihalf Embedded Systems
+sf-tec.de	Science Fiction Technologies
+sgi.com		SGI
+sicortex.com	Sicortex
+siemens.com	Siemens
+sierrawireless.com	Sierra Wireless
+sigma-chemnitz.de	SIGMA Chemnitz
+snapgear.com	Snapgear
+solidboot.com	Solid Boot Ltd.
+sony.co.jp	Sony
+sony.com	Sony
+sonycom.com	Sony
+spidernet.net	SpiderNet Services
+st.com		ST Microelectronics
+stlinux.com	ST Microelectronics
+starentnetworks.com	Starent Networks
+steeleye.com	SteelEye
+sun.com		Sun
+suse.com	Novell
+suse.cz		Novell
+suse.de		Novell
+sw.ru		Parallels
+swsoft.com	Parallels
+tapsys.com	Tapestry Systems
+telargo.com	Telargo
+tensilica.com	Tensilica
+terascala.com	Terascala
+thinktube.com	Thinktube
+ti.com		Texas Instruments
+til-technologies.fr	TIL Technologies
+tls.msk.ru	Telecom-Service
+toptica.com	TOPTICA Photonics
+toshiba.co.jp	Toshiba
+total-knowledge.com	Total Knowledge
+towertech.it	Tower Technologies
+tpi.com		TriplePoint
+transitive.com	Transitive
+transmode.se	Transmode Systems
+tresys.com	Tresys
+tripeaks.co.jp	Tripeaks
+trustedcs.com	Trusted Computer Solutions
+tungstengraphics.com	Tungsten Graphics
+tycho.nsa.gov	US National Security Agency
+ubuntu.com	Canonical
+uhulinux.hu	UHU-Linux
+unisys.com	Unisys
+valinux.co.jp	VA Linux Systems Japan
+verismonetworks.com	Verismo
+veritas.com	Veritas
+via.com.tw	Via
+vivecode.com	Vivecode
+vmware.com	VMWare
+volkswagen.de	Volkswagen
+voltaire.com	Voltaire
+vyatta.com	Vyatta
+wabtec.com	Wabtec Railway Electronics
+wacom.com	Wacom
+winbond.com	Winbond Electronics
+winbond.com.tw	Winbond Electronics
+wincor-nixdorf.com	Wincor Nixdorf
+windriver.com	Wind River
+wipro.com	Wipro
+wolfsonmicro.com	Wolfson Microelectronics
+xensource.com	XenSource
+xiv.co.il	XIV Information Systems
+xivstorage.com	XIV Information Systems
+trinnov.com	Trinnov Audio
+citrix.com	Citrix