-A (all_texts) option added for layout analysis

git-svn-id: https://pdfminerr.googlecode.com/svn/trunk/pdfminer@205 1aa58f4a-7d42-0410-adbc-911cccaed67c

pdfminer/converter.py

@@ -32,7 +32,8 @@ class PDFPageAggregator(PDFTextDevice):
     def end_page(self, _):
         assert not self.stack
         assert isinstance(self.cur_item, LTPage)
-        self.cur_item.fixate(self.laparams)
+        self.cur_item.fixate()
+        self.cur_item.analyze(self.laparams)
         self.pageno += 1
         return self.cur_item
 
@@ -43,7 +44,9 @@ class PDFPageAggregator(PDFTextDevice):
 
     def end_figure(self, _):
         fig = self.cur_item
+        assert isinstance(self.cur_item, LTFigure)
         self.cur_item.fixate()
+        self.cur_item.analyze(self.laparams)
         self.cur_item = self.stack.pop()
         self.cur_item.add(fig)
         return
@@ -226,14 +229,13 @@ class HTMLConverter(PDFConverter):
                 for child in item:
                     render(child)
             elif isinstance(item, LTImage):
-                name = ''
                 if self.outdir:
                     name = self.write_image(item)
-                self.outfp.write('<img src="%s" style="position:absolute; left:%dpx; top:%dpx;" '
+                    self.outfp.write('<img src="%s" style="position:absolute; left:%dpx; top:%dpx;" '
-                                 'width="%d" height="%d" />\n' %
+                                     'width="%d" height="%d" />\n' %
-                                 (enc(name),
+                                     (enc(name),
-                                  item.x0*self.scale, (self.yoffset-item.y1)*self.scale,
+                                      item.x0*self.scale, (self.yoffset-item.y1)*self.scale,
-                                  item.width*self.scale, item.height*self.scale))
+                                      item.width*self.scale, item.height*self.scale))
             return
         page = PDFConverter.end_page(self, page)
         render(page)
@@ -311,11 +313,13 @@ class XMLConverter(PDFConverter):
             elif isinstance(item, LTText):
                 self.outfp.write('<text>%s</text>\n' % item.text)
             elif isinstance(item, LTImage):
-                name = ''
                 if self.outdir:
                     name = self.write_image(item)
-                self.outfp.write('<image name="%s" width="%d" height="%d" />\n' %
+                    self.outfp.write('<image src="%s" width="%d" height="%d" />\n' %
-                                 (enc(name), item.width, item.height))
+                                     (enc(name), item.width, item.height))
+                else:
+                    self.outfp.write('<image width="%d" height="%d" />\n' %
+                                     (item.width, item.height))
             else:
                 assert 0, item
             return

pdfminer/layout.py

@@ -168,7 +168,7 @@ class LTImage(LTItem):
         self.bits = stream.get_any(('BPC', 'BitsPerComponent'), 1)
         self.colorspace = stream.get_any(('CS', 'ColorSpace'))
         if not isinstance(self.colorspace, list):
-            self.colorspace = [colorspace]
+            self.colorspace = [self.colorspace]
         return
 
     def __repr__(self):
@@ -550,41 +550,12 @@ def group_boxes(groupfunc, objs, distfunc, debug=0):
     return objs.pop()
 
 
-##  LTFigure
+##  LTAnalyzer
 ##
-class LTFigure(LTContainer):
+class LTAnalyzer(LTContainer):
 
-    def __init__(self, name, bbox, matrix):
+    def analyze(self, laparams):
-        (x,y,w,h) = bbox
-        bbox = get_bounds( apply_matrix_pt(matrix, (p,q))
-                           for (p,q) in ((x,y), (x+w,y), (x,y+h), (x+w,y+h)) )
-        self.name = name
-        self.matrix = matrix
-        LTContainer.__init__(self, bbox)
-        return
-
-    def __repr__(self):
-        return ('<figure %r bbox=%s matrix=%s>' %
-                (self.name, bbox2str(self.bbox), matrix2str(self.matrix)))
-
-
-##  LTPage
-##
-class LTPage(LTContainer):
-
-    def __init__(self, pageid, bbox, rotate=0):
-        LTContainer.__init__(self, bbox)
-        self.pageid = pageid
-        self.rotate = rotate
-        self.layout = None
-        return
-
-    def __repr__(self):
-        return ('<page(%r) bbox=%s rotate=%r>' % (self.pageid, bbox2str(self.bbox), self.rotate))
-
-    def fixate(self, laparams):
         """Perform the layout analysis."""
-        LTContainer.fixate(self)
         (textobjs, otherobjs) = self.get_textobjs()
         if not laparams or not textobjs: return
         if laparams.writing_mode not in ('lr-tb', 'tb-rl'):
@@ -694,3 +665,41 @@ class LTPage(LTContainer):
                     (max(obj1.y1,obj2.y1) - min(obj1.y0,obj2.y0)) -
                     (obj1.width*obj1.height + obj2.width*obj2.height))
         return group_boxes(LTTextGroupTBRL, boxes, dist)
+    
+
+##  LTFigure
+##
+class LTFigure(LTAnalyzer):
+
+    def __init__(self, name, bbox, matrix):
+        (x,y,w,h) = bbox
+        bbox = get_bounds( apply_matrix_pt(matrix, (p,q))
+                           for (p,q) in ((x,y), (x+w,y), (x,y+h), (x+w,y+h)) )
+        LTAnalyzer.__init__(self, bbox)
+        self.name = name
+        self.matrix = matrix
+        return
+
+    def __repr__(self):
+        return ('<figure %r bbox=%s matrix=%s>' %
+                (self.name, bbox2str(self.bbox), matrix2str(self.matrix)))
+
+    def analyze(self, laparams):
+        if laparams.all_texts:
+            LTAnalyzer.analyze(self, laparams)
+        return
+
+
+##  LTPage
+##
+class LTPage(LTAnalyzer):
+
+    def __init__(self, pageid, bbox, rotate=0):
+        LTAnalyzer.__init__(self, bbox)
+        self.pageid = pageid
+        self.rotate = rotate
+        self.layout = None
+        return
+
+    def __repr__(self):
+        return ('<page(%r) bbox=%s rotate=%r>' % (self.pageid, bbox2str(self.bbox), self.rotate))

tools/pdf2txt.py

@@ -12,11 +12,11 @@ def main(argv):
     import getopt
     def usage():
         print ('usage: %s [-d] [-p pagenos] [-m maxpages] [-P password] [-o output] '
-               '[-n] [-D writing_mode] [-M char_margin] [-L line_margin] [-W word_margin] '
+               '[-n] [-A] [-D writing_mode] [-M char_margin] [-L line_margin] [-W word_margin] '
                '[-O output_dir] [-t text|html|xml|tag] [-c codec] [-s scale] file ...' % argv[0])
         return 100
     try:
-        (opts, args) = getopt.getopt(argv[1:], 'dp:m:P:o:nD:M:L:W:O:t:c:s:')
+        (opts, args) = getopt.getopt(argv[1:], 'dp:m:P:o:nAD:M:L:W:O:t:c:s:')
     except getopt.GetoptError:
         return usage()
     if not args: return usage()
@@ -42,6 +42,7 @@ def main(argv):
         elif k == '-P': password = v
         elif k == '-o': outfile = v
         elif k == '-n': laparams = None
+        elif k == '-A': laparams.all_texts = True
         elif k == '-D': laparams.writing_mode = v
         elif k == '-M': laparams.char_margin = float(v)
         elif k == '-L': laparams.line_margin = float(v)