X-Git-Url: http://git.droids-corp.org/?p=diff2html.git;a=blobdiff_plain;f=diff2html.py;h=71ac19a3343b450c0d07a1a2683531caa9167129;hp=dc000eef6b1b12b8a1634864e4863adc2eed910a;hb=659c1a60bdf4fc73098bc97b48db70b68d82ad23;hpb=0ca475f9e41664da1195d4582567d52be5ab24d8
diff --git a/diff2html.py b/diff2html.py
index dc000ee..71ac19a 100644
--- a/diff2html.py
+++ b/diff2html.py
@@ -19,7 +19,7 @@
# Transform a unified diff from stdin to a colored
# side-by-side HTML page on stdout.
#
-# Authors: Olivier MATZ
+# Authors: Olivier Matz
# Alan De Smet
# Sergey Satskiy
# scito
@@ -29,13 +29,17 @@
#
# TODO:
# - The sane function currently mashes non-ASCII characters to "."
-# Instead be clever and convert to something like "xF0"
+# Instead be clever and convert to something like "xF0"
# (the hex value), and mark with a . Even more clever:
# Detect if the character is "printable" for whatever definition,
# and display those directly.
-
-import sys, re, htmlentitydefs, getopt, StringIO, codecs
+import sys, re, htmlentitydefs, getopt, StringIO, codecs, datetime
+try:
+ from simplediff import diff, string_diff
+except ImportError:
+ sys.stderr.write("info: simplediff module not found, only linediff is available\n")
+ sys.stderr.write("info: it can be downloaded at https://github.com/paulgb/simplediff\n")
# minimum line size, we add a zero-sized breakable space every
# LINESIZE characters
@@ -43,32 +47,49 @@ linesize = 20
tabsize = 8
show_CR = False
encoding = "utf-8"
+lang = "en"
+algorithm = 0
+desc = "File comparison"
+dtnow = datetime.datetime.now()
+modified_date = "%s+01:00"%dtnow.isoformat()
html_hdr = """
-
-
-
- HTML Diff
-
-
-
+
+
+
+
+
+ HTML Diff{0}
+
+
+
+
+
+
+
+
+
"""
html_footer = """
+
"""
@@ -90,7 +111,7 @@ hunk_off1, hunk_size1, hunk_off2, hunk_size2 = 0, 0, 0, 0
# Characters we're willing to word wrap on
-WORDBREAK = " \t;.,/):"
+WORDBREAK = " \t;.,/):-"
def sane(x):
r = ""
@@ -103,6 +124,9 @@ def sane(x):
return r
def linediff(s, t):
+ '''
+ Original line diff algorithm of diff2html. It's character based.
+ '''
if len(s):
s = unicode(reduce(lambda x, y:x+y, [ sane(c) for c in s ]))
if len(t):
@@ -165,6 +189,62 @@ def linediff(s, t):
return r1, r2
+def diff_changed(old, new):
+ '''
+ Returns the differences basend on characters between two strings
+ wrapped with DIFFON and DIFFOFF using `diff`.
+ '''
+ con = {'=': (lambda x: x),
+ '+': (lambda x: DIFFON + x + DIFFOFF),
+ '-': (lambda x: '')}
+ return "".join([(con[a])("".join(b)) for a, b in diff(old, new)])
+
+
+def diff_changed_ts(old, new):
+ '''
+ Returns a tuple for a two sided comparison based on characters, see `diff_changed`.
+ '''
+ return (diff_changed(new, old), diff_changed(old, new))
+
+
+def word_diff(old, new):
+ '''
+ Returns the difference between the old and new strings based on words. Punctuation is not part of the word.
+
+ Params:
+ old the old string
+ new the new string
+
+ Returns:
+ the output of `diff` on the two strings after splitting them
+ on whitespace (a list of change instructions; see the docstring
+ of `diff`)
+ '''
+ separator_pattern = '(\W+)';
+ return diff(re.split(separator_pattern, old, flags=re.UNICODE), re.split(separator_pattern, new, flags=re.UNICODE))
+
+
+def diff_changed_words(old, new):
+ '''
+ Returns the difference between two strings based on words (see `word_diff`)
+ wrapped with DIFFON and DIFFOFF.
+
+ Returns:
+ the output of the diff expressed delimited with DIFFON and DIFFOFF.
+ '''
+ con = {'=': (lambda x: x),
+ '+': (lambda x: DIFFON + x + DIFFOFF),
+ '-': (lambda x: '')}
+ return "".join([(con[a])("".join(b)) for a, b in word_diff(old, new)])
+
+
+def diff_changed_words_ts(old, new):
+ '''
+ Returns a tuple for a two sided comparison based on words, see `diff_changed_words`.
+ '''
+ return (diff_changed_words(new, old), diff_changed_words(old, new))
+
+
def convert(s, linesize=0, ponct=0):
i = 0
t = u""
@@ -227,6 +307,9 @@ def add_line(s1, s2, output_file):
global line1
global line2
+ orig1 = s1
+ orig2 = s2
+
if s1 == None and s2 == None:
type_name = "unmodified"
elif s1 == None or s1 == "":
@@ -237,7 +320,12 @@ def add_line(s1, s2, output_file):
type_name = "unmodified"
else:
type_name = "changed"
- s1, s2 = linediff(s1, s2)
+ if algorithm == 1:
+ s1, s2 = diff_changed_words_ts(orig1, orig2)
+ elif algorithm == 2:
+ s1, s2 = diff_changed_ts(orig1, orig2)
+ else: # default
+ s1, s2 = linediff(orig1, orig2)
output_file.write(('
' % type_name).encode(encoding))
if s1 != None and s1 != "":
@@ -295,14 +383,15 @@ def empty_buffer(output_file):
buf = []
-def parse_input(input_file, output_file,
+def parse_input(input_file, output_file, input_file_name, output_file_name,
exclude_headers, show_hunk_infos):
global add_cpt, del_cpt
global line1, line2
global hunk_off1, hunk_size1, hunk_off2, hunk_size2
if not exclude_headers:
- output_file.write(html_hdr.format(encoding).encode(encoding))
+ title_suffix = ' ' + input_file_name
+ output_file.write(html_hdr.format(title_suffix, encoding, desc, "", modified_date, lang).encode(encoding))
output_file.write(table_hdr.encode(encoding))
while True:
@@ -363,7 +452,7 @@ def parse_input(input_file, output_file,
empty_buffer(output_file)
output_file.write(table_footer.encode(encoding))
if not exclude_headers:
- output_file.write(html_footer.encode(encoding))
+ output_file.write(html_footer.format("", dtnow.strftime("%d.%m.%Y")).encode(encoding))
def usage():
@@ -383,6 +472,7 @@ stdout may not work with UTF-8, instead use -o option.
-l linesize set maximum line size is there is no word break (default 20)
-r show \\r characters
-k show hunk infos
+ -a algo line diff algorithm (0: linediff characters, 1: word, 2: simplediff characters) (default 0)
-h show help and exit
'''
@@ -390,18 +480,19 @@ def main():
global linesize, tabsize
global show_CR
global encoding
+ global algorithm
- input_file = sys.stdin
- output_file = sys.stdout
+ input_file_name = ''
+ output_file_name = ''
exclude_headers = False
show_hunk_infos = False
try:
- opts, args = getopt.getopt(sys.argv[1:], "he:i:o:xt:l:rk",
+ opts, args = getopt.getopt(sys.argv[1:], "he:i:o:xt:l:rka:",
["help", "encoding=", "input=", "output=",
"exclude-html-headers", "tabsize=",
- "linesize=", "show-cr", "show-hunk-infos"])
+ "linesize=", "show-cr", "show-hunk-infos", "algorithm="])
except getopt.GetoptError, err:
print unicode(err) # will print something like "option -a not recognized"
usage()
@@ -415,8 +506,10 @@ def main():
encoding = a
elif o in ("-i", "--input"):
input_file = codecs.open(a, "r", encoding)
+ input_file_name = a
elif o in ("-o", "--output"):
output_file = codecs.open(a, "w")
+ output_file_name = a
elif o in ("-x", "--exclude-html-headers"):
exclude_headers = True
elif o in ("-t", "--tabsize"):
@@ -427,16 +520,27 @@ def main():
show_CR = True
elif o in ("-k", "--show-hunk-infos"):
show_hunk_infos = True
+ elif o in ("-a", "--algorithm"):
+ algorithm = int(a)
else:
assert False, "unhandled option"
- parse_input(input_file, output_file,
+
+ # Use stdin if not input file is set
+ if not ('input_file' in locals()):
+ input_file = codecs.getreader(encoding)(sys.stdin)
+
+ # Use stdout if not output file is set
+ if not ('output_file' in locals()):
+ output_file = codecs.getwriter(encoding)(sys.stdout)
+
+ parse_input(input_file, output_file, input_file_name, output_file_name,
exclude_headers, show_hunk_infos)
def parse_from_memory(txt, exclude_headers, show_hunk_infos):
" Parses diff from memory and returns a string with html "
input_stream = StringIO.StringIO(txt)
output_stream = StringIO.StringIO()
- parse_input(input_stream, output_stream, exclude_headers, show_hunk_infos)
+ parse_input(input_stream, output_stream, '', '', exclude_headers, show_hunk_infos)
return output_stream.getvalue()