major: integrate ocr to pdf reader

backend/configs/config.py

@@ -4,6 +4,7 @@ from dotenv import load_dotenv
 basedir = os.path.abspath(os.path.dirname(__file__))
 load_dotenv(os.path.join(basedir, '.env'))
 
+os.environ["TESSDATA_PREFIX"] = os.path.join(basedir, 'tessdata')
 class Config(object):
     SECRET_KEY = os.environ.get('SECRET_KEY') or "s0m37h!n6-obfu5c471ng"
     SQLALCHEMY_DATABASE_URI = os.environ.get('DATABASE_URL', '').replace(

backend/configs/tessdata/configs/alto

@@ -0,0 +1 @@
+tessedit_create_alto 1

backend/configs/tessdata/configs/ambigs.train

@@ -0,0 +1,7 @@
+tessedit_ambigs_training	1
+load_freq_dawg	0
+load_punc_dawg	0
+load_system_dawg	0
+load_number_dawg	0
+ambigs_debug_level	3
+load_fixed_length_dawgs	0

backend/configs/tessdata/configs/api_config

@@ -0,0 +1 @@
+tessedit_zero_rejection T

backend/configs/tessdata/configs/bigram

@@ -0,0 +1,5 @@
+load_bigram_dawg	True
+tessedit_enable_bigram_correction	True
+tessedit_bigram_debug	3
+save_raw_choices	True
+save_alt_choices	True

backend/configs/tessdata/configs/box.train

@@ -0,0 +1,12 @@
+disable_character_fragments T
+file_type                   .bl
+textord_fast_pitch_test	T
+tessedit_zero_rejection T
+tessedit_minimal_rejection F
+tessedit_write_rep_codes F
+edges_children_fix F
+edges_childarea 0.65
+edges_boxarea 0.9
+tessedit_resegment_from_boxes T
+tessedit_train_from_boxes T
+textord_no_rejects T

backend/configs/tessdata/configs/box.train.stderr

@@ -0,0 +1,13 @@
+file_type .bl
+#tessedit_use_nn F
+textord_fast_pitch_test T
+tessedit_zero_rejection T
+tessedit_minimal_rejection F
+tessedit_write_rep_codes F
+edges_children_fix F
+edges_childarea 0.65
+edges_boxarea 0.9
+tessedit_resegment_from_boxes T
+tessedit_train_from_boxes T
+#textord_repeat_extraction F
+textord_no_rejects T

backend/configs/tessdata/configs/digits

@@ -0,0 +1 @@
+tessedit_char_whitelist 0123456789-.

backend/configs/tessdata/configs/get.images

@@ -0,0 +1 @@
+tessedit_write_images T

backend/configs/tessdata/configs/hocr

@@ -0,0 +1,2 @@
+tessedit_create_hocr 1
+hocr_font_info 0

backend/configs/tessdata/configs/inter

@@ -0,0 +1,2 @@
+interactive_display_mode				T
+tessedit_display_outwords		T

backend/configs/tessdata/configs/kannada

@@ -0,0 +1,4 @@
+textord_skewsmooth_offset 8
+textord_skewsmooth_offset2 8
+textord_merge_desc 0.5
+textord_no_rejects 1

backend/configs/tessdata/configs/linebox

@@ -0,0 +1,2 @@
+tessedit_resegment_from_line_boxes 1
+tessedit_make_boxes_from_boxes 1

backend/configs/tessdata/configs/logfile

@@ -0,0 +1 @@
+debug_file tesseract.log

backend/configs/tessdata/configs/lstm.train

@@ -0,0 +1,11 @@
+file_type                   .bl
+textord_fast_pitch_test	T
+tessedit_zero_rejection T
+tessedit_minimal_rejection F
+tessedit_write_rep_codes F
+edges_children_fix F
+edges_childarea 0.65
+edges_boxarea 0.9
+tessedit_train_line_recognizer T
+textord_no_rejects T
+tessedit_init_config_only T

backend/configs/tessdata/configs/lstmbox

@@ -0,0 +1 @@
+tessedit_create_lstmbox 1

backend/configs/tessdata/configs/lstmdebug

@@ -0,0 +1,4 @@
+stopper_debug_level 1
+classify_debug_level 1
+segsearch_debug_level 1
+language_model_debug_level 3

backend/configs/tessdata/configs/makebox

@@ -0,0 +1 @@
+tessedit_create_boxfile 1

backend/configs/tessdata/configs/pdf

@@ -0,0 +1 @@
+tessedit_create_pdf 1

backend/configs/tessdata/configs/quiet

@@ -0,0 +1 @@
+debug_file /dev/null

backend/configs/tessdata/configs/rebox

@@ -0,0 +1,2 @@
+tessedit_resegment_from_boxes 1
+tessedit_make_boxes_from_boxes 1

backend/configs/tessdata/configs/strokewidth

@@ -0,0 +1,12 @@
+textord_show_blobs 0
+textord_debug_tabfind 3
+textord_tabfind_show_partitions 1
+textord_tabfind_show_initial_partitions 1
+textord_tabfind_show_columns 1
+textord_tabfind_show_blocks 1
+textord_tabfind_show_initialtabs 1
+textord_tabfind_show_finaltabs 1
+textord_tabfind_show_strokewidths 1
+textord_tabfind_show_vlines 0
+textord_tabfind_show_images 1
+tessedit_dump_pageseg_images 0

backend/configs/tessdata/configs/tsv

@@ -0,0 +1 @@
+tessedit_create_tsv 1

backend/configs/tessdata/configs/txt

@@ -0,0 +1,3 @@
+# This config file should be used with other config files which create renderers.
+# usage example: tesseract eurotext.tif eurotext txt hocr pdf
+tessedit_create_txt 1

backend/configs/tessdata/configs/unlv

@@ -0,0 +1,2 @@
+tessedit_write_unlv 1
+unlv_tilde_crunching T

backend/configs/tessdata/configs/wordstrbox

@@ -0,0 +1 @@
+tessedit_create_wordstrbox 1

backend/configs/tessdata/eng.user-patterns

@@ -0,0 +1,2 @@
+1-\d\d\d-GOOG-411
+www.\n\\\*.com

backend/configs/tessdata/eng.user-words

@@ -0,0 +1,5 @@
+the
+quick
+brown
+fox
+jumped

backend/configs/tessdata/tessconfigs/batch

@@ -0,0 +1 @@
+# No content needed as all defaults are correct.

backend/configs/tessdata/tessconfigs/batch.nochop

@@ -0,0 +1,2 @@
+chop_enable 0
+wordrec_enable_assoc 0

backend/configs/tessdata/tessconfigs/matdemo

@@ -0,0 +1,7 @@
+#################################################
+# Adaptive Matcher Using PreAdapted Templates
+#################################################
+
+classify_enable_adaptive_debugger   1
+matcher_debug_flags         6
+matcher_debug_level       1

backend/configs/tessdata/tessconfigs/msdemo

@@ -0,0 +1,12 @@
+#################################################
+# Adaptive Matcher Using PreAdapted Templates
+#################################################
+
+classify_enable_adaptive_debugger	1
+matcher_debug_flags         6
+matcher_debug_level       1
+
+wordrec_display_splits          0
+wordrec_display_all_blobs       1
+wordrec_display_segmentations   2
+classify_debug_level		1

backend/configs/tessdata/tessconfigs/nobatch

@@ -0,0 +1 @@
+

backend/configs/tessdata/tessconfigs/segdemo

@@ -0,0 +1,9 @@
+#################################################
+# Adaptive Matcher Using PreAdapted Templates
+#################################################
+
+wordrec_display_splits          0
+wordrec_display_all_blobs       1
+wordrec_display_segmentations   2
+classify_debug_level		1
+stopper_debug_level 1

backend/migrations/versions/926395732c3e_raise_bonid_digits.py

@@ -0,0 +1,38 @@
+"""raise bonid digits
+
+Revision ID: 926395732c3e
+Revises: 2a64d3b9235a
+Create Date: 2024-08-24 10:33:39.109944
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = '926395732c3e'
+down_revision = '2a64d3b9235a'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('receipt', schema=None) as batch_op:
+        batch_op.alter_column('bonid',
+               existing_type=sa.NUMERIC(precision=24, scale=0),
+               type_=sa.Numeric(precision=28, scale=0),
+               existing_nullable=True)
+
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('receipt', schema=None) as batch_op:
+        batch_op.alter_column('bonid',
+               existing_type=sa.Numeric(precision=28, scale=0),
+               type_=sa.NUMERIC(precision=24, scale=0),
+               existing_nullable=True)
+
+    # ### end Alembic commands ###

backend/models/receipt.py

@@ -4,7 +4,7 @@ from src import db
 class Receipt(db.Model):
     id = db.Column(db.BigInteger, primary_key=True, autoincrement=True)
     date = db.Column(db.Date, nullable=False)
-    bonid = db.Column(db.Numeric(precision=24, scale=0), unique=True)
+    bonid = db.Column(db.Numeric(precision=28, scale=0), unique=True)
     from_user = db.Column(db.ForeignKey("login_token.token"),
                           server_onupdate=db.FetchedValue())
     registered = db.Column(db.Boolean, nullable=False,

backend/src/receipts/upload/routes.py

@@ -40,7 +40,7 @@ def upload_receipt(establishment: int):
             db.session.commit()
             if pdfReceipt:
                 rename(f"{PDFDir}/temp.pdf", f"{PDFDir}{secure_filename(f'{dbReceipt.id}.pdf')}")
-                LOGGER.debug(receipt.text)
+                LOGGER.debug(receipt.words)
             return redirect(url_for("receipts.check_items.confirm_receipt_items", receipt_id = dbReceipt.id))
         else:
             LOGGER.debug(form.errors)

backend/src/utils/modules/receipt_parser/edeka/edeka_parser.py

@@ -0,0 +1,20 @@
+from datetime import datetime
+
+def getDictFromWords(words: list[tuple]):
+    results = {"items": []}
+    results['date'] = datetime.strptime(words[-6][4], "%d.%m.%y").date()
+    results['bonid'] = words[-1][4]
+    currentline = 0
+    skipwords = 14
+    for i, word in enumerate(words[skipwords:]):
+        if currentline != word[5]:
+            results['items'].append({"itemname": word[4]})
+            currentline = word[5]
+        elif word[6] == 0:
+            results['items'][-1]["itemname"] += " " + word[4]
+        if word[6] == 1 and word[7] == 1:
+            results['items'][-1]["price"] = word[4].split("*")[0]
+        if "----" in word[4]:
+            del(results['items'][-1])
+            break
+    return results

backend/src/utils/modules/receipt_parser/kaufland/kaufland_parser.py

@@ -0,0 +1,27 @@
+from datetime import datetime
+
+def getDictFromWords(words: list[tuple]):
+    results = {"items": []}
+    results['bonid'] = words[-1][4]
+    currentline = 0
+    skipwords = 9
+    for i, word in enumerate(words[skipwords:]):
+        if currentline != word[5]:
+            results['items'].append({"itemname": word[4]})
+            currentline = word[5]
+        elif word[6] == 0:
+            results['items'][-1]["itemname"] += " " + word[4]
+        if word[6] == 1 and word[7] == 0:
+            if word[4].lower() == "x":
+                results['items'][-1]["amount"] = words[i+skipwords+1][4]
+            else:
+                results['items'][-1]["price"] = word[4]
+        elif word[6] == 2:
+                results['items'][-1]["price"] = word[4]
+        if word[4].lower() == "gesamt":
+             del(results['items'][-1])
+             break
+    for i, word in enumerate(words[::-1]):
+         if word[4].lower() == "datum:":
+            results['date'] = datetime.strptime(words[::-1][i-1][4], "%d.%m.%Y").date()
+    return results

backend/src/utils/modules/receipt_parser/pdf_receipt_parser.py

@@ -1,5 +1,7 @@
 import fitz
 from datetime import datetime, date
+from .edeka.edeka_parser import getDictFromWords as edekaparser
+from .kaufland.kaufland_parser import getDictFromWords as kauflandparser
 from re import search
 
 class PDFReceipt:
@@ -10,22 +12,29 @@ class PDFReceipt:
     parser -- A keyword in lowercase to tell how the receipt is formated.
         Currently supported: 'edeka'
     """
-    def __init__(self, bPDFFile, parser: str = "edeka") -> None:
+    def __init__(self, strPDFFile) -> None:
         try:
-            self.text = PDFReceipt._getTextFromPDF(bPDFFile)
-            self.id, self.date, self.items = PDFReceipt._getInfosFromText(self.text, parser)
+            self.words = PDFReceipt._getWordsFromPDF(strPDFFile)
+            storename = PDFReceipt._getStoreName(self.words)
+            self.id, self.date, self.items = PDFReceipt._getInfosFromText(self.words, store = storename)
         except:
-            self.text = "PDF konnte nicht geladen werden."
+            self.words = "PDF konnte nicht geladen werden."
             self.date = date.today()
             self.id = None
             self.items = []
 
-    def _getTextFromPDF(file):
+    def _getWordsFromPDF(file):
         with fitz.open(file, filetype="pdf") as doc:
-            text = ""
+            words = []
             for page in doc:
-                text += page.get_text()
-        return text.strip()
+                words.extend(page.get_text("words", textpage=page.get_textpage_ocr(), sort=True))
+        return words
+    
+    def _getStoreName(words: list[tuple]) -> str:
+        for word in words:
+            if word[4].lower() in ("edeka", "kaufland"):
+                return word[4].lower()
+        return "unknown"
 
     def _getItemsTextFromText(text, start="", end=""):
         return text[text.index(start)+len(start):text.index(end)].strip()
@@ -43,21 +52,23 @@ class PDFReceipt:
                 i += 2
         return resultsArr
 
-    def _getInfosFromText(text: str, parser: str = "edeka"):
-        if parser.lower() == "edeka":
-            items = PDFReceipt._convertItemsTextToDict(PDFReceipt._getItemsTextFromText(text, start="EUR", end="----------"))
-            strDate = text.split("\n")[-1].split(" ")[0]
-            date = datetime.strptime(strDate, "%d.%m.%y").date()
-            strReceiptNumber = text.split("\n")[-1].split(" ")[-1]
-            try:
-                intReceiptNumber = int(strReceiptNumber)
-            except:
-                raise ValueError("Receipt Number not an integer.")
+    def _getInfosFromText(words: str, store: str = "edeka"):
+        if store == "edeka":
+            result = edekaparser(words)
+        elif store == "kaufland":
+            result = kauflandparser(words)
+        items = result.get("items")
+        date = result.get("date")
+        strReceiptNumber = result.get("bonid")
+        try:
+            intReceiptNumber = int(strReceiptNumber)
+        except:
+            raise ValueError("Receipt Number not an integer.")
         return (intReceiptNumber, date, items)
 
-    def getPDFReceiptFromFile(strPDFFile: str, parser: str = "edeka"):
+    def getPDFReceiptFromFile(strPDFFile: str):
         try:
             with open(strPDFFile) as doc:
-                return PDFReceipt(doc, parser)
+                return PDFReceipt(doc)
         except FileNotFoundError as e:
             return PDFReceipt(None)