ag/pixelocr
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Use backend-specific suffixes for output filenames.

Browse source
This commit is contained in:
Andrey Golovizin 2014-09-08 21:15:55 +02:00
parent 721de38920
commit 4548f92e73
2 changed files with 4 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
pixelocr/formatting.py
View file

@ -21,6 +21,7 @@ from .glyphdb import WhitespaceData
class OutputFormat(object): class OutputFormat(object):
suffix = ''
last_tag = None last_tag = None
def __call__(self, glyph_data_stream): def __call__(self, glyph_data_stream):
@ -65,6 +66,7 @@ class OutputFormat(object):
class TextFormat(OutputFormat): class TextFormat(OutputFormat):
suffix = '.txt'
def assign_tag(self, glyph_data): def assign_tag(self, glyph_data):
return None return None
@ -73,6 +75,7 @@ class TextFormat(OutputFormat):
class HTMLFormat(OutputFormat): class HTMLFormat(OutputFormat):
suffix = '.html'
def assign_glyph_tag(self, glyph_data): def assign_glyph_tag(self, glyph_data):
style = glyph_data.style style = glyph_data.style
if style.bold: if style.bold:

2
pixelocr/gui/ocrengine.py
View file

@ -65,7 +65,7 @@ class OCREngine(QThread):
self.pageChanged.emit(page) self.pageChanged.emit(page)
page_text = self.recognize_page(page) page_text = self.recognize_page(page)
print(page_text) print(page_text)
with open(filename + '.txt', 'w') as page_text_file: with open(filename + self.output_format.suffix, 'w') as page_text_file:
page_text_file.write(page_text) page_text_file.write(page_text)
def recognize_page(self, page): def recognize_page(self, page):