From 035bc05637a351cd29f20c29f76589d5af3796f8 Mon Sep 17 00:00:00 2001
From: augustin64 <me.git@augustin64.fr>
Date: Fri, 11 Oct 2024 15:28:27 +0200
Subject: [PATCH] Initial commit

---
 .gitignore       |  3 ++
 README.md        |  3 ++
 cemantix.py      | 82 ++++++++++++++++++++++++++++++++++++++++++++++++
 requirements.txt |  2 ++
 4 files changed, 90 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 README.md
 create mode 100644 cemantix.py
 create mode 100644 requirements.txt

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..a7e2360
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,3 @@
+.venv
+**/__pycache__
+*.bin
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..32085dc
--- /dev/null
+++ b/README.md
@@ -0,0 +1,3 @@
+# [Cemantix](https://cemantix.certitudes.org/) CLI (fully local)
+
+Download models at [Jean-Philippe Fauconnier's website](https://fauconnier.github.io/#data) (`frWac_non_lem_no_postag_no_phrase_200_skip_cut100.bin` recommanded)
diff --git a/cemantix.py b/cemantix.py
new file mode 100644
index 0000000..37e5037
--- /dev/null
+++ b/cemantix.py
@@ -0,0 +1,82 @@
+from gensim.models import KeyedVectors
+from colorama import Fore, Back, Style
+import argparse
+import random
+
+
+def cemantix(model, word=None):
+    if word is None:
+        rd = random.randint(0, len(model))
+        word = model.index_to_key[rd]
+        while "_" in word:
+            rd += 1
+            word = model.index_to_key[rd]
+
+    nearest = [i[0] for i in model.most_similar(word, topn=1000)]
+    def get_rank(guess):
+        if guess not in nearest:
+            return None
+        return 1000 - nearest.index(guess)
+
+    def formatted_status(guesses, last=None):
+        text = ""
+        for w, temp, rank in guesses:
+            if rank is not None:
+                text += Back.RED+Fore.BLACK+Style.BRIGHT+str(rank)+Style.RESET_ALL
+            text += "\t"
+            if temp >= 75:
+                text += Fore.BLUE
+            text += str(temp) + Style.RESET_ALL + "\t"
+            if last == w:
+                text += Style.BRIGHT+Back.WHITE+Fore.BLACK
+            text += w + Style.RESET_ALL+"\n"
+        return text[:-1]
+
+    def tried(word, guessed):
+        return word in [i[0] for i in guessed]
+
+    guesses = []
+
+    while True:
+        try:
+            guess = input(Style.BRIGHT+"Your guess > "+Style.RESET_ALL).strip()
+        except (EOFError, KeyboardInterrupt):
+            print("The word was "+Style.BRIGHT+word+Style.RESET_ALL)
+            print("Goodbye!")
+            return -1
+        try:
+            dist = round(round(model.distance(word, guess), 4)*100, 2)
+        except KeyError:
+            print(Fore.RED+"Key not present"+Style.RESET_ALL)
+            continue
+        
+        if not tried(guess, guesses):
+            guesses.append((guess, dist, get_rank(guess)))
+            guesses.sort(key=lambda x:-x[1])
+        print(chr(27) + "[2J")
+        print(formatted_status(guesses, last=guess))
+        if guess == word:
+            print(Fore.GREEN+"Correct!"+Style.RESET_ALL+f" {len(guesses)} tries.")
+            return len(guesses)
+
+
+
+
+def main():
+    parser = argparse.ArgumentParser(description="local infinite cemantix")
+    parser.add_argument("-w", "--word", dest="word", action="store",
+                        help="Specify goal word")
+    parser.add_argument("-m", "--model", dest="model", action="store",
+                        default="frWac_non_lem_no_postag_no_phrase_200_skip_cut100.bin",
+                        help="Specify model to use")
+    args = parser.parse_args()
+
+    model = KeyedVectors.load_word2vec_format(
+        args.model,
+        binary=True,
+        unicode_errors="ignore"
+    )
+    cemantix(model, word=args.word)
+
+if __name__ == "__main__":
+    main()
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..dbab0f3
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,2 @@
+colorama
+gensim