diff options
| author | Guido van Rossum <guido@python.org> | 1995-09-22 00:54:32 (GMT) |
|---|---|---|
| committer | Guido van Rossum <guido@python.org> | 1995-09-22 00:54:32 (GMT) |
| commit | 3c0bfd0deeadccf10c8982d5b42dd39f63fb727e (patch) | |
| tree | b818dcb9a68ec84ea0819b0e850c306a4cbf0da3 | |
| parent | 8421c4e833dc8bc70c5e98ee872e3b914e8ccda9 (diff) | |
| download | cpython-3c0bfd0deeadccf10c8982d5b42dd39f63fb727e.zip cpython-3c0bfd0deeadccf10c8982d5b42dd39f63fb727e.tar.gz cpython-3c0bfd0deeadccf10c8982d5b42dd39f63fb727e.tar.bz2 | |
fix <!...!> parsing; added verbose option; don't lowercase entityrefs
| -rw-r--r-- | Lib/sgmllib.py | 12 |
1 files changed, 7 insertions, 5 deletions
diff --git a/Lib/sgmllib.py b/Lib/sgmllib.py index 0169ea3..fc023eb 100644 --- a/Lib/sgmllib.py +++ b/Lib/sgmllib.py @@ -40,7 +40,8 @@ commentclose = regex.compile('--[ \t\n]*>') class SGMLParser: # Interface -- initialize and reset this instance - def __init__(self): + def __init__(self, verbose=0): + self.verbose = verbose self.reset() # Interface -- reset this instance. Loses all unprocessed data @@ -141,7 +142,8 @@ class SGMLParser: k = incomplete.match(rawdata, i) if k < 0: raise RuntimeError, 'no incomplete match ??' j = i+k - if j == n: break # Really incomplete + if j == n or rawdata[i:i+2] == '<!': + break # Really incomplete self.handle_data(rawdata[i:j]) i = j # end while @@ -234,8 +236,9 @@ class SGMLParser: # Example -- report an unbalanced </...> tag. def report_unbalanced(self, tag): - print '*** Unbalanced </' + tag + '>' - print '*** Stack:', self.stack + if self.verbose: + print '*** Unbalanced </' + tag + '>' + print '*** Stack:', self.stack # Example -- handle character reference, no need to override def handle_charref(self, name): @@ -256,7 +259,6 @@ class SGMLParser: # Example -- handle entity reference, no need to override def handle_entityref(self, name): table = self.entitydefs - name = string.lower(name) if table.has_key(name): self.handle_data(table[name]) else: |