gitignore update + run.py reformat

.gitignore

@@ -6,3 +6,6 @@
 *.o
 .DS_Store
 .token
+scripts/*
+solution/*
+models/*

run.py

@@ -1,24 +1,18 @@
 #!/usr/bin/python3
-import sys
+from scripts.language_utils import predict_gaps, train_model, load_lexicon
 
 
-for line in sys.stdin:
-    _, _, _, _, _, _, left_context, right_context = line.split("\t")
+WORDS_LIMIT = 10
 
-    previous_word = left_context.split()[-1]
-    next_word = right_context.split()[0]
-    
-    if previous_word == "United" and next_word == "of":
-        print("States:0.9 :0.1")
-    elif previous_word == "used":
-        print("to:0.4 it:0.3 as:0.2 :0.1")
-    elif previous_word.lower() == "in":
-        print("the:0.7 a:0.1 an:0.1 :0.1")
-    elif previous_word.lower() == "i":
-        print("am:0.3 was:0.3 have:0.3 :0.1")
-    elif previous_word.lower() in ["he", "she", "it"]:
-        print("is:0.3 was:0.3 has:0.3 :0.1")
-    elif previous_word.lower() in "bring":
-        print("something:0.3 it:0.3 them:0.3 :0.1")
-    else:
-        print("the:0.5 a:0.2 an:0.2 :0.1")
+if __name__ == "__main__":
+    lexicon = load_lexicon("./words_alpha.txt")
+    model = train_model("./train/in.tsv.xz", "./train/expected.tsv", lexicon)
+    for i, (directory, lines_no) in enumerate(zip(["dev-0", "test-A"], (10519, 7414))):
+        predict_gaps(
+            model,
+            f"./{directory}/in.tsv.xz",
+            f"./{directory}/out.tsv",
+            lexicon,
+            lines_no,
+            WORDS_LIMIT,
+        )