Check output before processing the whole file for ISBN and DOI search

README.md

@@ -108,8 +108,6 @@ A list of ideas and TODO. Don't hesitate to give feedback on the ones you really
 10. Refactor
     11. Use bibtex-parser lib to write bibtex, instead of parsed2BibTex
     12. Rebuild function
-    15. Check output of subprocesses before it ends
-    16. TODO in files
 20. No DOI for arXiv / HAL
 30. Parameter to disable remote search
 40. Open file

fetcher.py

@@ -50,19 +50,26 @@ def findISBN(src):
     if src.endswith(".pdf"):
         totext = subprocess.Popen(["pdftotext", src, "-"],
                                   stdout=subprocess.PIPE,
-                                  stderr=subprocess.PIPE)
+                                  stderr=subprocess.PIPE,
+                                  bufsize=1)
     elif src.endswith(".djvu"):
         totext = subprocess.Popen(["djvutxt", src],
                                   stdout=subprocess.PIPE,
-                                  stderr=subprocess.PIPE)
-    extractfull = totext.communicate()
-    # TODO : ^ Return result before processing the whole book ?
-    if extractfull[1] is not "":
-        # Error happened
-        tools.warning(extractfull[1])
-        return False
-    extractfull = extractfull[0]
+                                  stderr=subprocess.PIPE,
+                                  bufsize=1)
+    while totext.poll() == None:
+        extractfull = totext.stdin.readline()
         extractISBN = isbn_re.search(extractfull.lower().replace('Œ', '-'))
+        if extractISBN:
+            totext.terminate()
+            break
+
+    err = totext.communicate()[1]
+    if totext.returncode > 0:
+        # Error happened
+        tools.warning(err)
+        return False
+
     cleanISBN = False
     # Clean ISBN is the ISBN number without separators
     if extractISBN:
@@ -103,13 +110,9 @@ def findDOI(src):
         totext = subprocess.Popen(["djvutxt", src],
                                   stdout=subprocess.PIPE,
                                   stderr=subprocess.PIPE)
-    extractfull = totext.communicate()
-    # TODO : ^ Return result before full conversion ?
-    if extractfull[1] is not "":
-        # Error happened
-        tools.warning(extractfull[1])
-        return False
-    extractfull = extractfull[0]
+
+    while totext.poll() == None:
+        extractfull = totext.stdin.readline()
         extractDOI = doi_re.search(extractfull.lower().replace('Œ', '-'))
         if not extractDOI:
             # PNAS fix
@@ -118,6 +121,15 @@ def findDOI(src):
             if not extractDOI:
                 # JSB fix
                 extractDOI = doi_jsb_re.search(extractfull.lower())
+        if extractDOI:
+            totext.terminate()
+            break
+
+    err = totext.communicate()[1]
+    if totext.returncode > 0:
+        # Error happened
+        tools.warning(err)
+        return False
 
     cleanDOI = False
     if extractDOI:

tearpages.py

@@ -52,7 +52,7 @@ def tearpage(filename):
 
     # Write pages excepted the first one
     output_file = PdfFileWriter()
-    for i in range(0, num_pages):
+    for i in range(1, num_pages):
         output_file.addPage(input_file.getPage(i))
 
     tmp.close()