Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -17,10 +17,14 @@ from pipelines.scoring import compute_quality_score
|
|
| 17 |
from pipelines.storage import persist_to_hf
|
| 18 |
from pipelines.utils import detect_filetype, load_doc_text
|
| 19 |
|
| 20 |
-
|
|
|
|
|
|
|
| 21 |
|
|
|
|
| 22 |
|
| 23 |
def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
|
|
|
| 24 |
if not filepaths:
|
| 25 |
raise gr.Error("少なくとも1ファイルをアップロードしてください。")
|
| 26 |
|
|
@@ -31,6 +35,7 @@ def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
|
| 31 |
filename = os.path.basename(path)
|
| 32 |
with open(path, "rb") as f:
|
| 33 |
raw_bytes = f.read()
|
|
|
|
| 34 |
filetype = detect_filetype(filename, raw_bytes)
|
| 35 |
|
| 36 |
# 1) テキスト抽出
|
|
@@ -79,7 +84,7 @@ def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
|
| 79 |
# 8) 要約
|
| 80 |
summaries = summarize_with_openai(merged_text)
|
| 81 |
|
| 82 |
-
# 9)
|
| 83 |
result_json = {
|
| 84 |
"candidate_id": candidate_id or hashlib.sha256(merged_text.encode("utf-8")).hexdigest()[:16],
|
| 85 |
"files": [os.path.basename(p) for p in filepaths],
|
|
@@ -107,7 +112,6 @@ def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
|
| 107 |
|
| 108 |
anon_pdf = (result_json["candidate_id"] + ".anon.pdf", anon_pdf_bytes)
|
| 109 |
|
| 110 |
-
# 重要:gradioのAPIスキーマ生成バグ回避のため、dictはすべて文字列で返す
|
| 111 |
return (
|
| 112 |
json.dumps(result_json, ensure_ascii=False, indent=2), # out_json -> Code(str)
|
| 113 |
json.dumps(skills, ensure_ascii=False, indent=2), # out_skills -> Code(str)
|
|
@@ -115,7 +119,7 @@ def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
|
| 115 |
summaries["300chars"],
|
| 116 |
summaries["100chars"],
|
| 117 |
summaries["onesent"],
|
| 118 |
-
anon_pdf,
|
| 119 |
json.dumps(commit_info or {"status": "skipped (DATASET_REPO not set)"}, ensure_ascii=False, indent=2),
|
| 120 |
)
|
| 121 |
|
|
@@ -128,7 +132,7 @@ with gr.Blocks(title=APP_TITLE) as demo:
|
|
| 128 |
label="レジュメ類 (PDF/画像/Word/テキスト) 複数可",
|
| 129 |
file_count="multiple",
|
| 130 |
file_types=[".pdf", ".png", ".jpg", ".jpeg", ".tiff", ".bmp", ".docx", ".txt"],
|
| 131 |
-
type="filepath", # ← 'file'
|
| 132 |
)
|
| 133 |
candidate_id = gr.Textbox(label="候補者ID(任意。未入力なら自動生成)")
|
| 134 |
notes = gr.Textbox(label="補足メモ(任意)", lines=3)
|
|
@@ -136,13 +140,13 @@ with gr.Blocks(title=APP_TITLE) as demo:
|
|
| 136 |
run_btn = gr.Button("実行")
|
| 137 |
|
| 138 |
with gr.Tab("構造化JSON"):
|
| 139 |
-
out_json = gr.Code(label="統合出力 (JSON)")
|
| 140 |
|
| 141 |
with gr.Tab("抽出スキル"):
|
| 142 |
-
out_skills = gr.Code(label="スキル一覧 (JSON)") # ← gr.JSON
|
| 143 |
|
| 144 |
with gr.Tab("品質スコア"):
|
| 145 |
-
out_score = gr.Code(label="品質評価 (JSON)")
|
| 146 |
|
| 147 |
with gr.Tab("要約 (300/100/1文)"):
|
| 148 |
out_sum_300 = gr.Textbox(label="300字要約")
|
|
@@ -153,7 +157,7 @@ with gr.Blocks(title=APP_TITLE) as demo:
|
|
| 153 |
out_pdf = gr.File(label="匿名PDFダウンロード")
|
| 154 |
|
| 155 |
with gr.Tab("Datasets 保存ログ"):
|
| 156 |
-
out_commit = gr.Code(label="コミット情報")
|
| 157 |
|
| 158 |
run_btn.click(
|
| 159 |
process_resumes,
|
|
@@ -162,5 +166,6 @@ with gr.Blocks(title=APP_TITLE) as demo:
|
|
| 162 |
)
|
| 163 |
|
| 164 |
if __name__ == "__main__":
|
| 165 |
-
#
|
| 166 |
-
demo.launch(
|
|
|
|
|
|
| 17 |
from pipelines.storage import persist_to_hf
|
| 18 |
from pipelines.utils import detect_filetype, load_doc_text
|
| 19 |
|
| 20 |
+
# ❶ app.py の該当箇所を以下に置き換え
|
| 21 |
+
|
| 22 |
+
# ...(importsはそのまま)
|
| 23 |
|
| 24 |
+
APP_TITLE = "候補者インテーク & レジュメ標準化(OpenAI版)"
|
| 25 |
|
| 26 |
def process_resumes(filepaths, candidate_id: str, additional_notes: str = ""):
|
| 27 |
+
import os, json, hashlib, gradio as gr
|
| 28 |
if not filepaths:
|
| 29 |
raise gr.Error("少なくとも1ファイルをアップロードしてください。")
|
| 30 |
|
|
|
|
| 35 |
filename = os.path.basename(path)
|
| 36 |
with open(path, "rb") as f:
|
| 37 |
raw_bytes = f.read()
|
| 38 |
+
|
| 39 |
filetype = detect_filetype(filename, raw_bytes)
|
| 40 |
|
| 41 |
# 1) テキスト抽出
|
|
|
|
| 84 |
# 8) 要約
|
| 85 |
summaries = summarize_with_openai(merged_text)
|
| 86 |
|
| 87 |
+
# 9) 構造化出力(UIバグ回避のため **すべて文字列** で返す)
|
| 88 |
result_json = {
|
| 89 |
"candidate_id": candidate_id or hashlib.sha256(merged_text.encode("utf-8")).hexdigest()[:16],
|
| 90 |
"files": [os.path.basename(p) for p in filepaths],
|
|
|
|
| 112 |
|
| 113 |
anon_pdf = (result_json["candidate_id"] + ".anon.pdf", anon_pdf_bytes)
|
| 114 |
|
|
|
|
| 115 |
return (
|
| 116 |
json.dumps(result_json, ensure_ascii=False, indent=2), # out_json -> Code(str)
|
| 117 |
json.dumps(skills, ensure_ascii=False, indent=2), # out_skills -> Code(str)
|
|
|
|
| 119 |
summaries["300chars"],
|
| 120 |
summaries["100chars"],
|
| 121 |
summaries["onesent"],
|
| 122 |
+
anon_pdf,
|
| 123 |
json.dumps(commit_info or {"status": "skipped (DATASET_REPO not set)"}, ensure_ascii=False, indent=2),
|
| 124 |
)
|
| 125 |
|
|
|
|
| 132 |
label="レジュメ類 (PDF/画像/Word/テキスト) 複数可",
|
| 133 |
file_count="multiple",
|
| 134 |
file_types=[".pdf", ".png", ".jpg", ".jpeg", ".tiff", ".bmp", ".docx", ".txt"],
|
| 135 |
+
type="filepath", # ← ここが重要:'file' は非対応・例外の元
|
| 136 |
)
|
| 137 |
candidate_id = gr.Textbox(label="候補者ID(任意。未入力なら自動生成)")
|
| 138 |
notes = gr.Textbox(label="補足メモ(任意)", lines=3)
|
|
|
|
| 140 |
run_btn = gr.Button("実行")
|
| 141 |
|
| 142 |
with gr.Tab("構造化JSON"):
|
| 143 |
+
out_json = gr.Code(label="統合出力 (JSON文字列)")
|
| 144 |
|
| 145 |
with gr.Tab("抽出スキル"):
|
| 146 |
+
out_skills = gr.Code(label="スキル一覧 (JSON文字列)") # ← gr.JSON を **使わない**
|
| 147 |
|
| 148 |
with gr.Tab("品質スコア"):
|
| 149 |
+
out_score = gr.Code(label="品質評価 (JSON文字列)")
|
| 150 |
|
| 151 |
with gr.Tab("要約 (300/100/1文)"):
|
| 152 |
out_sum_300 = gr.Textbox(label="300字要約")
|
|
|
|
| 157 |
out_pdf = gr.File(label="匿名PDFダウンロード")
|
| 158 |
|
| 159 |
with gr.Tab("Datasets 保存ログ"):
|
| 160 |
+
out_commit = gr.Code(label="コミット情報 (JSON文字列)")
|
| 161 |
|
| 162 |
run_btn.click(
|
| 163 |
process_resumes,
|
|
|
|
| 166 |
)
|
| 167 |
|
| 168 |
if __name__ == "__main__":
|
| 169 |
+
# ローカル(gradio.live)想定なら share=True が安全
|
| 170 |
+
# Spaces 内では demo.launch() のままでもOK
|
| 171 |
+
demo.launch(share=True, server_name="0.0.0.0")
|