From 9151351ac9aa69f13a347474cfabb289fdacc1dc Mon Sep 17 00:00:00 2001
From: diegantobass <diego.antolinosbasso@sciencespo.fr>
Date: Wed, 7 May 2025 10:00:29 +0200
Subject: [PATCH] refine parsing with pythhonvCard4

---
 vcf_cleaner.py | 38 ++++++++++++++++++++++++++------------
 1 file changed, 26 insertions(+), 12 deletions(-)

diff --git a/vcf_cleaner.py b/vcf_cleaner.py
index 2164255..0e8aea0 100644
--- a/vcf_cleaner.py
+++ b/vcf_cleaner.py
@@ -1,19 +1,33 @@
 import pickle
 from pythonvCard4.vcard import Contact
+import base64
+import os
 
-input = open("contacts.vcf", 'r').readlines()
-output = open("processed.vcf", 'w')
+file = open("contacts.vcf", 'r').readlines()
 
-for line in input:
+cards = []
 
-    if "TEL" in line:
-        line = line.replace('-', '')
-        if ":06" in line: line = line.replace(":06", ":+336")
-        if ":07" in line: line = line.replace(":07", ":+337")
-        if ":00" in line: line = line.replace(":00", ":+")
-        if ":09" in line: line = line.replace(":09", ":+339")
-        if ":01" in line: line = line.replace(":01", ":+331")
+current_card = ""
+for line in file:
+    current_card += line
+    if "END:VCARD" in line:
+        contact = Contact.from_vcard(current_card)
+        if "PHOTO" in contact.custom and len(contact.custom["PHOTO"]) > 1:
+            os.makedirs("photo/" + contact.fn, exist_ok=True)
+            for image in range(len(contact.custom["PHOTO"])):
+                with open("photo/" + contact.fn + "/" + str(image) + ".jpg", "wb") as f:
+                    f.write(base64.decodebytes(str.encode(contact.custom["PHOTO"][image])))
 
-        print(line)
+        cards.append(contact)
+        current_card = ""
+        continue
 
-    output.write(line)
+
+# contact = Contact.from_vcard(cards[156])
+# print(contact.custom["PHOTO"])
+
+# vcf_text = contact.to_vcard()
+# print(vcf_text)
+# open('test.vcf', 'w').write(vcf_text)
+
+print(len(cards))