Upload 2 files

Touten.py

@@ -0,0 +1,79 @@
+
+import MeCab
+# 形態素解析器の初期化
+mecab = MeCab.Tagger()
+
+
+def check_end_type(line):
+    parsed = mecab.parse(line)
+    # print(parsed)
+    parts = parsed.split("\n")
+    final_part = parts[-3]
+    temp = final_part.split("\t")
+    end_type = temp[-2]
+    hinshi = temp[4]
+    return end_type, hinshi
+
+
+def touten_insert(text):
+    new_lines = []
+    old_line = ""
+    current_line = ""
+    non_kaiyo = False
+
+    for line in text.split("\n"):
+        if line == "":
+            continue
+        if line == old_line:
+            continue
+        old_line = line
+
+        if non_kaiyo:
+            current_line += line
+            # print("non_kaiyo added", line)
+            non_kaiyo = False
+            continue
+
+        end_type, hinshi = check_end_type(line)
+        if end_type.find("終止形") >= 0 or end_type.find("命令") >= 0:
+            current_line += line+"。"
+        elif hinshi.find("名詞") >= 0 or hinshi.find("記号"):
+            new_lines.append(current_line)
+            current_line = ""
+            new_lines.append(line)
+        else:
+            current_line += line
+    new_lines.append(current_line)
+    new_lines = [i for i in new_lines if i != ""]
+
+    return "\n".join(new_lines)
+
+
+def end_filter(line):
+    end_symbols = [
+        "、", "（", "("
+    ]
+    for symb in end_symbols:
+        if line.endswith(symb):
+            return True
+    return False
+
+
+def del_kaigyo(text):
+    new_lines = []
+    temp_line = ""
+    for line in text.split("\n"):
+        # print(line)
+        if end_filter(line) or end_filter(temp_line):
+            # print("integ:", temp_line, "----", line)
+            temp_line += line
+
+        else:
+            if temp_line != "":
+                new_lines.append(temp_line)
+            new_lines.append(line)
+            temp_line = ""
+
+        # if line.startswith("を") or line.startswith("は"):
+        #    current_line += line
+    return "\n".join(new_lines)

load_jap.py

@@ -0,0 +1,136 @@
+# %%
+# jap2010を読み込んでparquetにするcode｡実行不要
+
+"""
+テキストアーカイブ - 日本語ウェブコーパス 2010
+http://www.s-yata.jp/corpus/nwc2010/texts/
+を処理するcode｡
+
+#URLは常時は空いていないので注意｡
+wget http://www.s-yata.jp/corpus/nwc2010/texts/filelist
+wget -i filelist
+"""
+
+# %%
+from datasets import load_dataset
+import glob
+import pandas as pd
+import lzma
+from Touten import touten_insert, del_kaigyo
+
+# %%
+import os
+
+out_dir = "../data/jap2010_parquet"
+if not os.path.exists(out_dir):
+    os.mkdir(out_dir)
+
+# %%
+
+
+# %%
+count_threshold = 3
+max_lines = 2000
+
+
+task_name = "00000001"
+for xz_path in glob.glob("../data/jap2010/*.xz"):
+    task_name = xz_path.split("/")[-1].split(".")[0]
+    print(task_name)
+    path = f"../data/jap2010/{task_name}.xz"
+    # .xz ファイルを開いて内容をチャンクで読み込む
+    with lzma.open(path, 'rt') as file:
+        lines = file.readlines()
+
+    new_line_count = 0
+
+    current_doc = []
+    cnt = 0
+    text_len = 0
+    text_list = []
+    for line in (lines):
+        if line == "\n":
+            new_line_count += 1
+        else:
+            new_line_count = 0
+        current_doc.append(line)
+        if new_line_count >= count_threshold or len(current_doc) > max_lines:
+            text = "\n".join(current_doc)
+            text = text.strip()
+            text = text.replace("\n\n", "\n")
+            text = touten_insert(text)
+            text = del_kaigyo(text)
+            current_doc = []
+            new_line_count = 0
+
+            """
+            if text!="":
+                d=json.dumps({"text":text},ensure_ascii=False)
+                with open(f"{out_dir}/{task_name}.jsonl","a") as f:
+                    f.write(f"{d}\n")
+            """
+            if text != "":
+                # print(len(text))
+                text_list.append(text)
+                text_len += len(text)
+
+        if text_len > 10**8:
+            # if len(text_list)>10:
+            #    break
+
+            df = pd.DataFrame(text_list, columns=["text"])  # .reset_index()
+            df.to_parquet(f"{out_dir}/{task_name}_{cnt}.parquet")
+            cnt += 1
+            text_list = []
+            text_len = 0
+            # if cnt>100:
+            #    break
+    # break
+    df = pd.DataFrame(text_list, columns=["text"]).reset_index()
+    df.to_parquet(f"{out_dir}/{task_name}_{cnt}.parquet")
+
+# %%
+text = text_list[1]
+text = touten_insert(text)
+text = del_kaigyo(text)
+print(text)
+
+# %%
+text = del_kaigyo(text)
+print(text)
+
+
+# %%
+t = "たくさんの「わたし」に見てもらいたいから、ダムの映画祭を、谷根千で、"
+# end_filter(t)
+
+# %%
+df = pd.read_parquet("data/jap2010_parquet/00000001_0.parquet")
+df
+
+# %%
+# huggingface-cli upload --repo-type=dataset hatakeyama-llm-team/japanese2010 .
+
+# %%
+# load test
+dataset = load_dataset("hatakeyama-llm-team/japanese2010",
+                       streaming=True,
+                       split="train",
+                       )
+
+# %%
+cnt = 0
+for data in dataset:
+    print(data["text"][:10])
+    cnt += 1
+    if cnt > 10:
+        break
+
+# %%
+
+
+# %%
+with open("t.txt", "w") as f:
+    f.write(data["text"])
+
+# %%