Fix converting path to multiple rectangles

For path that consists of a series of rectangles
(shape is 'mlllhmlllh...'), call paint_path again with each group of
5 points. The result is multiple rects instead of a single curve.

fixes #369

pdfminer/converter.py

@@ -27,6 +27,9 @@
 
 
 class PDFLayoutAnalyzer(PDFTextDevice):
+
+    RECTS = re.compile('^(mlllh)+$')
+
     def __init__(self, rsrcmgr, pageno=1, laparams=None):
         PDFTextDevice.__init__(self, rsrcmgr)
         self.pageno = pageno
@@ -100,6 +103,12 @@ def paint_path(self, gstate, stroke, fill, evenodd, path):
                                          stroke, fill, evenodd, gstate.scolor,
                                          gstate.ncolor))
                 return
+        if self.RECTS.match(shape):
+            for path_group in zip(*(iter(path),) * 5):
+                self.paint_path(gstate, stroke, fill, evenodd,
+                                list(path_group))
+            return
+
         # other shapes
         pts = []
         for p in path:

tests/test_tools_pdf2txt.py

@@ -71,6 +71,9 @@ def test_contrib_hash_two_complement(self):
         """
         run('contrib/issue-00352-hash-twos-complement.pdf')
 
+    def test_contrib_excel(self):
+        run('contrib/excel.pdf', '-t xml')
+
 
 class TestDumpImages: