all-in-mcp 0.1.4__tar.gz → 0.2.2__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/PKG-INFO +13 -2
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/README.md +12 -1
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/docs/api.md +55 -0
- all_in_mcp-0.2.2/docs/cryptobib.md +166 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/pyproject.toml +1 -1
- all_in_mcp-0.2.2/src/all_in_mcp/academic_platforms/__init__.py +6 -0
- all_in_mcp-0.2.2/src/all_in_mcp/academic_platforms/cryptobib.py +461 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/src/all_in_mcp/server.py +139 -2
- all_in_mcp-0.1.4/src/all_in_mcp/academic_platforms/__init__.py +0 -1
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/.gitignore +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/LICENSE +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/docs/configuration.md +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/docs/development.md +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/docs/iacr.md +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/docs/pypi-setup.md +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/src/all_in_mcp/__init__.py +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/src/all_in_mcp/academic_platforms/base.py +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/src/all_in_mcp/academic_platforms/iacr.py +0 -0
- {all_in_mcp-0.1.4 → all_in_mcp-0.2.2}/src/all_in_mcp/paper.py +0 -0
@@ -1,6 +1,6 @@
|
|
1
1
|
Metadata-Version: 2.4
|
2
2
|
Name: all-in-mcp
|
3
|
-
Version: 0.
|
3
|
+
Version: 0.2.2
|
4
4
|
Summary: An MCP (Model Context Protocol) server providing daily-use utility functions and academic paper search capabilities
|
5
5
|
Project-URL: Homepage, https://github.com/jiahaoxiang2000/all-in-mcp
|
6
6
|
Project-URL: Repository, https://github.com/jiahaoxiang2000/all-in-mcp
|
@@ -52,14 +52,24 @@ An MCP (Model Context Protocol) server that provides daily-use utility functions
|
|
52
52
|
### Daily Utilities
|
53
53
|
|
54
54
|
- **Academic Research**: IACR ePrint Archive paper search, download, and reading
|
55
|
+
- **Bibliography Search**: CryptoBib database search for cryptography papers
|
55
56
|
|
56
57
|
### Paper Search Capabilities
|
57
58
|
|
59
|
+
#### IACR ePrint Archive
|
60
|
+
|
58
61
|
- Search academic papers from IACR ePrint Archive
|
59
62
|
- Download PDF files
|
60
63
|
- Extract and read text content from papers
|
61
64
|
- Metadata extraction (authors, publication dates, abstracts)
|
62
65
|
|
66
|
+
#### CryptoBib Database
|
67
|
+
|
68
|
+
- Search comprehensive cryptography bibliography database
|
69
|
+
- Access to thousands of cryptographic research papers
|
70
|
+
- Retrieve structured paper metadata or raw BibTeX entries
|
71
|
+
- Support for all major cryptography venues and conferences
|
72
|
+
|
63
73
|
## Quick Start
|
64
74
|
|
65
75
|
### Prerequisites
|
@@ -75,7 +85,7 @@ An MCP (Model Context Protocol) server that provides daily-use utility functions
|
|
75
85
|
pip install all-in-mcp
|
76
86
|
```
|
77
87
|
|
78
|
-
###
|
88
|
+
### Option 2: Install from Source
|
79
89
|
|
80
90
|
1. Clone this repository:
|
81
91
|
|
@@ -201,6 +211,7 @@ Complete documentation is available in the [`docs/`](docs/) directory:
|
|
201
211
|
- **[API Reference](docs/api.md)** - Complete API documentation
|
202
212
|
- **[Installation Guide](docs/installation.md)** - Setup instructions
|
203
213
|
- **[IACR Integration](docs/iacr.md)** - Academic paper search details
|
214
|
+
- **[CryptoBib Integration](docs/cryptobib.md)** - Bibliography database search
|
204
215
|
- **[Development Guide](docs/development.md)** - Contributing guidelines
|
205
216
|
- **[PyPI Setup Guide](docs/pypi-setup.md)** - Publishing configuration
|
206
217
|
- **[Examples](docs/examples.md)** - Usage examples
|
@@ -7,14 +7,24 @@ An MCP (Model Context Protocol) server that provides daily-use utility functions
|
|
7
7
|
### Daily Utilities
|
8
8
|
|
9
9
|
- **Academic Research**: IACR ePrint Archive paper search, download, and reading
|
10
|
+
- **Bibliography Search**: CryptoBib database search for cryptography papers
|
10
11
|
|
11
12
|
### Paper Search Capabilities
|
12
13
|
|
14
|
+
#### IACR ePrint Archive
|
15
|
+
|
13
16
|
- Search academic papers from IACR ePrint Archive
|
14
17
|
- Download PDF files
|
15
18
|
- Extract and read text content from papers
|
16
19
|
- Metadata extraction (authors, publication dates, abstracts)
|
17
20
|
|
21
|
+
#### CryptoBib Database
|
22
|
+
|
23
|
+
- Search comprehensive cryptography bibliography database
|
24
|
+
- Access to thousands of cryptographic research papers
|
25
|
+
- Retrieve structured paper metadata or raw BibTeX entries
|
26
|
+
- Support for all major cryptography venues and conferences
|
27
|
+
|
18
28
|
## Quick Start
|
19
29
|
|
20
30
|
### Prerequisites
|
@@ -30,7 +40,7 @@ An MCP (Model Context Protocol) server that provides daily-use utility functions
|
|
30
40
|
pip install all-in-mcp
|
31
41
|
```
|
32
42
|
|
33
|
-
###
|
43
|
+
### Option 2: Install from Source
|
34
44
|
|
35
45
|
1. Clone this repository:
|
36
46
|
|
@@ -156,6 +166,7 @@ Complete documentation is available in the [`docs/`](docs/) directory:
|
|
156
166
|
- **[API Reference](docs/api.md)** - Complete API documentation
|
157
167
|
- **[Installation Guide](docs/installation.md)** - Setup instructions
|
158
168
|
- **[IACR Integration](docs/iacr.md)** - Academic paper search details
|
169
|
+
- **[CryptoBib Integration](docs/cryptobib.md)** - Bibliography database search
|
159
170
|
- **[Development Guide](docs/development.md)** - Contributing guidelines
|
160
171
|
- **[PyPI Setup Guide](docs/pypi-setup.md)** - Publishing configuration
|
161
172
|
- **[Examples](docs/examples.md)** - Usage examples
|
@@ -108,6 +108,61 @@ URL: https://eprint.iacr.org/2023/1234
|
|
108
108
|
[Full extracted text content]
|
109
109
|
```
|
110
110
|
|
111
|
+
## CryptoBib Bibliography Search
|
112
|
+
|
113
|
+
### search-cryptobib-papers
|
114
|
+
|
115
|
+
Search CryptoBib bibliography database for cryptography papers. CryptoBib is a comprehensive BibTeX database of cryptography-related academic papers.
|
116
|
+
|
117
|
+
**Parameters:**
|
118
|
+
|
119
|
+
- `query` (string, required): Search query string (e.g., 'cryptography', 'lattice', 'homomorphic')
|
120
|
+
- `max_results` (integer, optional): Maximum number of results to return (default: 10)
|
121
|
+
- `return_bibtex` (boolean, optional): Whether to return raw BibTeX entries (default: false)
|
122
|
+
- `force_download` (boolean, optional): Force download the newest crypto.bib file (default: false)
|
123
|
+
|
124
|
+
**Returns:**
|
125
|
+
|
126
|
+
- List of papers with metadata or raw BibTeX entries
|
127
|
+
|
128
|
+
**Example:**
|
129
|
+
|
130
|
+
```json
|
131
|
+
{
|
132
|
+
"name": "search-cryptobib-papers",
|
133
|
+
"arguments": {
|
134
|
+
"query": "lattice cryptography",
|
135
|
+
"max_results": 5,
|
136
|
+
"return_bibtex": true,
|
137
|
+
"force_download": false
|
138
|
+
}
|
139
|
+
}
|
140
|
+
```
|
141
|
+
|
142
|
+
**Response:**
|
143
|
+
|
144
|
+
````text
|
145
|
+
Found 5 BibTeX entries for query 'lattice cryptography':
|
146
|
+
|
147
|
+
Entry 1:
|
148
|
+
```bibtex
|
149
|
+
@InProceedings{CRYPTO:RegLyr05,
|
150
|
+
author = "Oded Regev and Ricky Steinfeld",
|
151
|
+
title = "Learning With Errors over Rings",
|
152
|
+
booktitle = "CRYPTO 2005",
|
153
|
+
year = 2005,
|
154
|
+
...
|
155
|
+
}
|
156
|
+
````
|
157
|
+
|
158
|
+
**Important Notes:**
|
159
|
+
|
160
|
+
- **First Use**: On first use, the tool will download the crypto.bib file (~50MB) and cache it locally in the `./downloads` directory
|
161
|
+
- **Caching**: Subsequent searches use the cached file unless `force_download` is set to true
|
162
|
+
- **BibTeX Retrieval**: To get specific BibTeX entries, use the `search-cryptobib-papers` tool with `return_bibtex=true` and a specific search query
|
163
|
+
- **Content**: The crypto.bib file contains thousands of cryptography-related papers from major conferences and journals
|
164
|
+
- **Performance**: Local file searching is much faster than streaming searches
|
165
|
+
|
111
166
|
## Error Handling
|
112
167
|
|
113
168
|
All tools return error messages in case of failures:
|
@@ -0,0 +1,166 @@
|
|
1
|
+
# CryptoBib Integration
|
2
|
+
|
3
|
+
This document describes the CryptoBib integration in the all-in-mcp server.
|
4
|
+
|
5
|
+
## Overview
|
6
|
+
|
7
|
+
CryptoBib (https://cryptobib.di.ens.fr/) is a comprehensive bibliography database for cryptography research papers. This integration allows searching through the CryptoBib database and retrieving BibTeX entries.
|
8
|
+
|
9
|
+
## Features
|
10
|
+
|
11
|
+
### Search Functionality
|
12
|
+
|
13
|
+
- Search through the entire CryptoBib database
|
14
|
+
- Case-insensitive search across all BibTeX fields
|
15
|
+
- Returns both parsed Paper objects and raw BibTeX entries
|
16
|
+
- Configurable result limits
|
17
|
+
|
18
|
+
### BibTeX Support
|
19
|
+
|
20
|
+
- Parse BibTeX entries into structured Paper objects
|
21
|
+
- Retrieve raw BibTeX entries for direct use
|
22
|
+
- Support for all standard BibTeX fields (author, title, booktitle, journal, year, etc.)
|
23
|
+
- Clean LaTeX formatting from titles and authors
|
24
|
+
|
25
|
+
## Available Tools
|
26
|
+
|
27
|
+
### search-cryptobib-papers
|
28
|
+
|
29
|
+
Search CryptoBib bibliography database for cryptography papers.
|
30
|
+
|
31
|
+
**Parameters:**
|
32
|
+
|
33
|
+
- `query` (required): Search query string (e.g., 'lattice', 'homomorphic encryption')
|
34
|
+
- `max_results` (optional, default: 10): Maximum number of results to return
|
35
|
+
- `return_bibtex` (optional, default: false): Whether to return raw BibTeX entries
|
36
|
+
- `force_download` (optional, default: false): Force download the newest crypto.bib file
|
37
|
+
|
38
|
+
**Examples:**
|
39
|
+
|
40
|
+
```json
|
41
|
+
{
|
42
|
+
"query": "lattice cryptography",
|
43
|
+
"max_results": 5,
|
44
|
+
"return_bibtex": false
|
45
|
+
}
|
46
|
+
```
|
47
|
+
|
48
|
+
To get specific BibTeX entries, you can use targeted search queries with `return_bibtex=true`.
|
49
|
+
|
50
|
+
## Usage Examples
|
51
|
+
|
52
|
+
### Basic Search
|
53
|
+
|
54
|
+
```python
|
55
|
+
from all_in_mcp.academic_platforms.cryptobib import CryptoBibSearcher
|
56
|
+
|
57
|
+
searcher = CryptoBibSearcher()
|
58
|
+
|
59
|
+
# Search for papers
|
60
|
+
papers = searcher.search("homomorphic encryption", max_results=5)
|
61
|
+
for paper in papers:
|
62
|
+
print(f"Title: {paper.title}")
|
63
|
+
print(f"Authors: {', '.join(paper.authors)}")
|
64
|
+
print(f"BibTeX Key: {paper.paper_id}")
|
65
|
+
```
|
66
|
+
|
67
|
+
### Get Raw BibTeX
|
68
|
+
|
69
|
+
```python
|
70
|
+
# Get raw BibTeX entries
|
71
|
+
bibtex_entries = searcher.search_bibtex("zero knowledge", max_results=3)
|
72
|
+
for entry in bibtex_entries:
|
73
|
+
print(entry)
|
74
|
+
```
|
75
|
+
|
76
|
+
### Get Specific BibTeX Entry
|
77
|
+
|
78
|
+
```python
|
79
|
+
# Search for specific entries using targeted queries
|
80
|
+
papers = searcher.search("ACISP:LZXSW24", max_results=1)
|
81
|
+
if papers:
|
82
|
+
# Access the BibTeX from the paper's extra field
|
83
|
+
bibtex = papers[0].extra.get('bibtex', '')
|
84
|
+
print(bibtex)
|
85
|
+
|
86
|
+
# Or get raw BibTeX entries directly
|
87
|
+
bibtex_entries = searcher.search_bibtex("ACISP:LZXSW24", max_results=1)
|
88
|
+
if bibtex_entries:
|
89
|
+
print(bibtex_entries[0])
|
90
|
+
```
|
91
|
+
|
92
|
+
## BibTeX Entry Format
|
93
|
+
|
94
|
+
CryptoBib uses standard BibTeX format. Example entry:
|
95
|
+
|
96
|
+
```bibtex
|
97
|
+
@InProceedings{ACISP:LZXSW24,
|
98
|
+
author = "Fangzhou Liu and
|
99
|
+
Xueqi Zhu and
|
100
|
+
Ruozhou Xu and
|
101
|
+
Danping Shi and
|
102
|
+
Peng Wang",
|
103
|
+
title = "The Offline Quantum Attack Against Modular Addition Variant of {Even}-{Mansour} Cipher",
|
104
|
+
pages = "3--19",
|
105
|
+
editor = acisp24ed,
|
106
|
+
booktitle = acisp24name1,
|
107
|
+
volume = acisp24vol1,
|
108
|
+
address = acisp24addr,
|
109
|
+
month = acisp24month,
|
110
|
+
publisher = acisp24pub,
|
111
|
+
series = mylncs,
|
112
|
+
year = 2024,
|
113
|
+
doi = "10.1007/978-981-97-5025-2_1",
|
114
|
+
}
|
115
|
+
```
|
116
|
+
|
117
|
+
## Implementation Details
|
118
|
+
|
119
|
+
### Search Strategy
|
120
|
+
|
121
|
+
The implementation uses streaming HTTP requests to search through the large crypto.bib file without downloading it entirely. It:
|
122
|
+
|
123
|
+
1. Makes a streaming request to the crypto.bib file
|
124
|
+
2. Parses BibTeX entries on-the-fly
|
125
|
+
3. Performs case-insensitive matching against the search query
|
126
|
+
4. Returns results as they are found up to the specified limit
|
127
|
+
|
128
|
+
### Performance Considerations
|
129
|
+
|
130
|
+
- Uses streaming to avoid memory issues with the large bibliography file
|
131
|
+
- Processes entries incrementally
|
132
|
+
- Stops searching once the maximum number of results is reached
|
133
|
+
- Implements proper brace counting for correct BibTeX entry parsing
|
134
|
+
|
135
|
+
### Limitations
|
136
|
+
|
137
|
+
- No PDF download capability (CryptoBib is a bibliography database only)
|
138
|
+
- No full-text search of paper content
|
139
|
+
- Search is limited to BibTeX field content
|
140
|
+
- Dependent on CryptoBib server availability
|
141
|
+
|
142
|
+
## Testing
|
143
|
+
|
144
|
+
Run the CryptoBib tests:
|
145
|
+
|
146
|
+
```bash
|
147
|
+
# Run all tests
|
148
|
+
pytest tests/test_cryptobib.py
|
149
|
+
|
150
|
+
# Run with verbose output
|
151
|
+
pytest tests/test_cryptobib.py -v
|
152
|
+
|
153
|
+
# Test specific functionality
|
154
|
+
python test_cryptobib.py
|
155
|
+
```
|
156
|
+
|
157
|
+
## Error Handling
|
158
|
+
|
159
|
+
The implementation includes comprehensive error handling for:
|
160
|
+
|
161
|
+
- Network connectivity issues
|
162
|
+
- Malformed BibTeX entries
|
163
|
+
- Server response errors
|
164
|
+
- Parsing failures
|
165
|
+
|
166
|
+
All errors are logged and gracefully handled, returning empty results or error messages as appropriate.
|
@@ -0,0 +1,461 @@
|
|
1
|
+
# all_in_mcp/academic_platforms/cryptobib.py
|
2
|
+
import logging
|
3
|
+
import os
|
4
|
+
import random
|
5
|
+
import re
|
6
|
+
from datetime import datetime
|
7
|
+
from pathlib import Path
|
8
|
+
from typing import ClassVar
|
9
|
+
|
10
|
+
import requests
|
11
|
+
|
12
|
+
from ..paper import Paper
|
13
|
+
from .base import PaperSource
|
14
|
+
|
15
|
+
logger = logging.getLogger(__name__)
|
16
|
+
|
17
|
+
|
18
|
+
class CryptoBibSearcher(PaperSource):
|
19
|
+
"""CryptoBib (https://cryptobib.di.ens.fr/) bibliography search implementation"""
|
20
|
+
|
21
|
+
CRYPTOBIB_BASE_URL = "https://cryptobib.di.ens.fr"
|
22
|
+
CRYPTOBIB_BIB_URL = "https://cryptobib.di.ens.fr/cryptobib/static/files/crypto.bib"
|
23
|
+
BROWSERS: ClassVar[list[str]] = [
|
24
|
+
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
|
25
|
+
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7)",
|
26
|
+
"Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36",
|
27
|
+
]
|
28
|
+
|
29
|
+
def __init__(self, cache_dir: str = "./downloads"):
|
30
|
+
self.cache_dir = Path(cache_dir)
|
31
|
+
self.cache_dir.mkdir(exist_ok=True)
|
32
|
+
self.bib_file_path = self.cache_dir / "crypto.bib"
|
33
|
+
self._setup_session()
|
34
|
+
|
35
|
+
def _setup_session(self):
|
36
|
+
"""Initialize session with random user agent"""
|
37
|
+
self.session = requests.Session()
|
38
|
+
self.session.headers.update(
|
39
|
+
{
|
40
|
+
"User-Agent": random.choice(self.BROWSERS),
|
41
|
+
"Accept": "text/plain,text/html,application/xhtml+xml",
|
42
|
+
"Accept-Language": "en-US,en;q=0.9",
|
43
|
+
}
|
44
|
+
)
|
45
|
+
|
46
|
+
def _download_bib_file(self, force_download: bool = False) -> bool:
|
47
|
+
"""
|
48
|
+
Download the crypto.bib file if not exists or if force_download is True
|
49
|
+
|
50
|
+
Args:
|
51
|
+
force_download: Force download even if file exists
|
52
|
+
|
53
|
+
Returns:
|
54
|
+
bool: True if file is ready, False if download failed
|
55
|
+
"""
|
56
|
+
try:
|
57
|
+
# Check if file exists and force_download is False
|
58
|
+
if self.bib_file_path.exists() and not force_download:
|
59
|
+
logger.info(f"Using cached crypto.bib file at {self.bib_file_path}")
|
60
|
+
return True
|
61
|
+
|
62
|
+
logger.info("Downloading crypto.bib file from CryptoBib...")
|
63
|
+
response = self.session.get(self.CRYPTOBIB_BIB_URL, stream=True)
|
64
|
+
|
65
|
+
if response.status_code != 200:
|
66
|
+
logger.error(
|
67
|
+
f"Failed to download crypto.bib: HTTP {response.status_code}"
|
68
|
+
)
|
69
|
+
return False
|
70
|
+
|
71
|
+
# Download with progress indication
|
72
|
+
total_size = int(response.headers.get("content-length", 0))
|
73
|
+
downloaded = 0
|
74
|
+
|
75
|
+
with open(self.bib_file_path, "wb") as f:
|
76
|
+
for chunk in response.iter_content(
|
77
|
+
chunk_size=1024 * 1024
|
78
|
+
): # 1MB chunks
|
79
|
+
if chunk:
|
80
|
+
f.write(chunk)
|
81
|
+
downloaded += len(chunk)
|
82
|
+
if total_size > 0:
|
83
|
+
progress = (downloaded / total_size) * 100
|
84
|
+
if downloaded % (1024 * 1024 * 5) == 0: # Log every 5MB
|
85
|
+
logger.info(f"Download progress: {progress:.1f}%")
|
86
|
+
|
87
|
+
logger.info(f"Successfully downloaded crypto.bib to {self.bib_file_path}")
|
88
|
+
return True
|
89
|
+
|
90
|
+
except requests.RequestException as e:
|
91
|
+
logger.error(f"Error downloading crypto.bib: {e}")
|
92
|
+
return False
|
93
|
+
except Exception as e:
|
94
|
+
logger.error(f"Unexpected error downloading crypto.bib: {e}")
|
95
|
+
return False
|
96
|
+
|
97
|
+
def _parse_bibtex_entry(self, bibtex_text: str) -> Paper | None:
|
98
|
+
"""Parse a single BibTeX entry into a Paper object"""
|
99
|
+
try:
|
100
|
+
# Extract entry type and key
|
101
|
+
entry_match = re.match(r"@(\w+){([^,]+),", bibtex_text, re.IGNORECASE)
|
102
|
+
if not entry_match:
|
103
|
+
return None
|
104
|
+
|
105
|
+
entry_type = entry_match.group(1).lower()
|
106
|
+
entry_key = entry_match.group(2).strip()
|
107
|
+
|
108
|
+
# Initialize fields
|
109
|
+
title = ""
|
110
|
+
authors = []
|
111
|
+
year = None
|
112
|
+
booktitle = ""
|
113
|
+
journal = ""
|
114
|
+
pages = ""
|
115
|
+
doi = ""
|
116
|
+
url = ""
|
117
|
+
abstract = ""
|
118
|
+
|
119
|
+
# Extract fields using a more robust approach
|
120
|
+
# First, normalize the text by removing extra whitespace
|
121
|
+
normalized_text = re.sub(r"\s+", " ", bibtex_text)
|
122
|
+
|
123
|
+
# Extract fields with better pattern matching
|
124
|
+
field_dict = {}
|
125
|
+
|
126
|
+
# Pattern for quoted fields (handles multi-line)
|
127
|
+
quoted_pattern = r'(\w+)\s*=\s*"([^"]*(?:[^"\\]|\\.)*)"'
|
128
|
+
for match in re.finditer(quoted_pattern, bibtex_text, re.DOTALL):
|
129
|
+
field_name = match.group(1).lower().strip()
|
130
|
+
field_value = match.group(2).strip()
|
131
|
+
field_dict[field_name] = field_value
|
132
|
+
|
133
|
+
# Pattern for unquoted fields (like numbers)
|
134
|
+
unquoted_pattern = r'(\w+)\s*=\s*([^,}\n"]+)'
|
135
|
+
for match in re.finditer(unquoted_pattern, bibtex_text):
|
136
|
+
field_name = match.group(1).lower().strip()
|
137
|
+
field_value = match.group(2).strip()
|
138
|
+
# Only add if not already present from quoted pattern
|
139
|
+
if field_name not in field_dict:
|
140
|
+
field_dict[field_name] = field_value
|
141
|
+
|
142
|
+
for field_name, field_value in field_dict.items():
|
143
|
+
field_name = field_name.lower().strip()
|
144
|
+
field_value = field_value.strip()
|
145
|
+
|
146
|
+
if field_name == "title":
|
147
|
+
# Clean up title (remove LaTeX commands)
|
148
|
+
title = re.sub(r"[{}]", "", field_value)
|
149
|
+
title = re.sub(r"\\[a-zA-Z]+", "", title)
|
150
|
+
title = title.strip()
|
151
|
+
|
152
|
+
elif field_name == "author":
|
153
|
+
# Parse authors - handle "and" separator
|
154
|
+
author_text = re.sub(r"[{}]", "", field_value)
|
155
|
+
authors = [
|
156
|
+
author.strip()
|
157
|
+
for author in re.split(r"\s+and\s+", author_text)
|
158
|
+
if author.strip()
|
159
|
+
]
|
160
|
+
|
161
|
+
elif field_name == "year":
|
162
|
+
try:
|
163
|
+
year = int(field_value)
|
164
|
+
except ValueError:
|
165
|
+
pass
|
166
|
+
|
167
|
+
elif field_name == "booktitle":
|
168
|
+
booktitle = re.sub(r"[{}]", "", field_value)
|
169
|
+
|
170
|
+
elif field_name == "journal":
|
171
|
+
journal = re.sub(r"[{}]", "", field_value)
|
172
|
+
|
173
|
+
elif field_name == "pages":
|
174
|
+
pages = field_value
|
175
|
+
|
176
|
+
elif field_name == "doi":
|
177
|
+
doi = field_value
|
178
|
+
|
179
|
+
elif field_name == "url":
|
180
|
+
url = field_value
|
181
|
+
|
182
|
+
elif field_name == "abstract":
|
183
|
+
abstract = field_value
|
184
|
+
|
185
|
+
# Determine venue (journal or conference)
|
186
|
+
venue = journal if journal else booktitle
|
187
|
+
categories = [entry_type] if entry_type else []
|
188
|
+
|
189
|
+
# Create published date
|
190
|
+
published_date = datetime(year, 1, 1) if year else datetime(1900, 1, 1)
|
191
|
+
|
192
|
+
return Paper(
|
193
|
+
paper_id=entry_key,
|
194
|
+
title=title,
|
195
|
+
authors=authors,
|
196
|
+
abstract=abstract,
|
197
|
+
url=url,
|
198
|
+
pdf_url="", # CryptoBib doesn't provide PDF URLs
|
199
|
+
published_date=published_date,
|
200
|
+
updated_date=None,
|
201
|
+
source="cryptobib",
|
202
|
+
categories=categories,
|
203
|
+
keywords=[],
|
204
|
+
doi=doi,
|
205
|
+
citations=0,
|
206
|
+
extra={
|
207
|
+
"bibtex": bibtex_text.strip(),
|
208
|
+
"venue": venue,
|
209
|
+
"pages": pages,
|
210
|
+
"entry_type": entry_type,
|
211
|
+
},
|
212
|
+
)
|
213
|
+
|
214
|
+
except Exception as e:
|
215
|
+
logger.warning(f"Failed to parse BibTeX entry: {e}")
|
216
|
+
return None
|
217
|
+
|
218
|
+
def search_bibtex(
|
219
|
+
self,
|
220
|
+
query: str,
|
221
|
+
max_results: int = 10,
|
222
|
+
force_download: bool = False,
|
223
|
+
year_min: int | None = None,
|
224
|
+
year_max: int | None = None,
|
225
|
+
) -> list[str]:
|
226
|
+
"""
|
227
|
+
Search CryptoBib and return raw BibTeX entries
|
228
|
+
|
229
|
+
Args:
|
230
|
+
query: Search query string
|
231
|
+
max_results: Maximum number of results to return
|
232
|
+
force_download: Force download the newest crypto.bib file
|
233
|
+
year_min: Minimum publication year (inclusive, optional)
|
234
|
+
year_max: Maximum publication year (inclusive, optional)
|
235
|
+
|
236
|
+
Returns:
|
237
|
+
List[str]: List of BibTeX entries as strings
|
238
|
+
|
239
|
+
Example:
|
240
|
+
>>> searcher = CryptoBibSearcher()
|
241
|
+
>>> # Search for recent implementation papers (2020-2024)
|
242
|
+
>>> entries = searcher.search_bibtex("implement", max_results=5,
|
243
|
+
... year_min=2020, year_max=2024)
|
244
|
+
>>>
|
245
|
+
>>> # Search for older RSA papers (1980-2000)
|
246
|
+
>>> entries = searcher.search_bibtex("RSA", max_results=10,
|
247
|
+
... year_min=1980, year_max=2000)
|
248
|
+
"""
|
249
|
+
bibtex_entries = []
|
250
|
+
|
251
|
+
try:
|
252
|
+
# Ensure we have the bib file locally
|
253
|
+
if not self._download_bib_file(force_download=force_download):
|
254
|
+
logger.error("Failed to download crypto.bib file")
|
255
|
+
return bibtex_entries
|
256
|
+
|
257
|
+
# Search in the local file
|
258
|
+
logger.info(f"Searching local crypto.bib file for: {query}")
|
259
|
+
current_entry = ""
|
260
|
+
in_entry = False
|
261
|
+
brace_count = 0
|
262
|
+
|
263
|
+
# Convert query to lowercase for case-insensitive search
|
264
|
+
query_lower = query.lower()
|
265
|
+
|
266
|
+
with open(self.bib_file_path, "r", encoding="utf-8") as f:
|
267
|
+
for line_num, line in enumerate(f, 1):
|
268
|
+
# Check if this is the start of a new entry
|
269
|
+
if line.strip().startswith("@") and not in_entry:
|
270
|
+
current_entry = line
|
271
|
+
in_entry = True
|
272
|
+
brace_count = line.count("{") - line.count("}")
|
273
|
+
elif in_entry:
|
274
|
+
current_entry += line
|
275
|
+
brace_count += line.count("{") - line.count("}")
|
276
|
+
|
277
|
+
# Check if entry is complete
|
278
|
+
if brace_count <= 0:
|
279
|
+
# Entry is complete, check if it matches the query
|
280
|
+
if query_lower in current_entry.lower():
|
281
|
+
# Check year range if specified
|
282
|
+
if self._entry_matches_year_range(
|
283
|
+
current_entry, year_min, year_max
|
284
|
+
):
|
285
|
+
bibtex_entries.append(current_entry.strip())
|
286
|
+
logger.info(
|
287
|
+
f"Found matching entry {len(bibtex_entries)} at line {line_num}"
|
288
|
+
)
|
289
|
+
|
290
|
+
if len(bibtex_entries) >= max_results:
|
291
|
+
break
|
292
|
+
|
293
|
+
# Reset for next entry
|
294
|
+
current_entry = ""
|
295
|
+
in_entry = False
|
296
|
+
brace_count = 0
|
297
|
+
|
298
|
+
except FileNotFoundError:
|
299
|
+
logger.error(f"crypto.bib file not found at {self.bib_file_path}")
|
300
|
+
except Exception as e:
|
301
|
+
logger.error(f"CryptoBib search error: {e}")
|
302
|
+
|
303
|
+
return bibtex_entries[:max_results]
|
304
|
+
|
305
|
+
def search(
|
306
|
+
self,
|
307
|
+
query: str,
|
308
|
+
max_results: int = 10,
|
309
|
+
return_bibtex: bool = False,
|
310
|
+
force_download: bool = False,
|
311
|
+
year_min: int | None = None,
|
312
|
+
year_max: int | None = None,
|
313
|
+
) -> list[Paper]:
|
314
|
+
"""
|
315
|
+
Search CryptoBib bibliography
|
316
|
+
|
317
|
+
Args:
|
318
|
+
query: Search query string
|
319
|
+
max_results: Maximum number of results to return
|
320
|
+
return_bibtex: If True, include raw BibTeX in results
|
321
|
+
force_download: Force download the newest crypto.bib file
|
322
|
+
year_min: Minimum publication year (inclusive, optional)
|
323
|
+
year_max: Maximum publication year (inclusive, optional)
|
324
|
+
|
325
|
+
Returns:
|
326
|
+
List[Paper]: List of paper objects
|
327
|
+
|
328
|
+
Example:
|
329
|
+
>>> searcher = CryptoBibSearcher()
|
330
|
+
>>> # Search for recent zero-knowledge papers (2020-2024)
|
331
|
+
>>> papers = searcher.search("zero knowledge", max_results=5,
|
332
|
+
... year_min=2020, year_max=2024)
|
333
|
+
>>>
|
334
|
+
>>> # Search for classic RSA papers (1977-1990)
|
335
|
+
>>> papers = searcher.search("RSA", max_results=10,
|
336
|
+
... year_min=1977, year_max=1990)
|
337
|
+
"""
|
338
|
+
papers = []
|
339
|
+
|
340
|
+
try:
|
341
|
+
# Get BibTeX entries
|
342
|
+
bibtex_entries = self.search_bibtex(
|
343
|
+
query,
|
344
|
+
max_results,
|
345
|
+
force_download=force_download,
|
346
|
+
year_min=year_min,
|
347
|
+
year_max=year_max,
|
348
|
+
)
|
349
|
+
|
350
|
+
# Parse each entry into Paper objects
|
351
|
+
for i, bibtex_text in enumerate(bibtex_entries):
|
352
|
+
logger.info(f"Parsing entry {i+1}/{len(bibtex_entries)}")
|
353
|
+
paper = self._parse_bibtex_entry(bibtex_text)
|
354
|
+
if paper:
|
355
|
+
papers.append(paper)
|
356
|
+
|
357
|
+
except Exception as e:
|
358
|
+
logger.error(f"CryptoBib search error: {e}")
|
359
|
+
|
360
|
+
return papers
|
361
|
+
|
362
|
+
def download_pdf(self, paper_id: str, save_path: str) -> str:
|
363
|
+
"""
|
364
|
+
CryptoBib doesn't provide PDF downloads
|
365
|
+
"""
|
366
|
+
return "Error: CryptoBib is a bibliography database and doesn't provide PDF downloads"
|
367
|
+
|
368
|
+
def read_paper(self, paper_id: str, save_path: str = "./downloads") -> str:
|
369
|
+
"""
|
370
|
+
CryptoBib doesn't provide paper content reading
|
371
|
+
"""
|
372
|
+
return "Error: CryptoBib is a bibliography database and doesn't provide paper content"
|
373
|
+
|
374
|
+
def get_bibtex_by_key(
|
375
|
+
self, entry_key: str, force_download: bool = False
|
376
|
+
) -> str | None:
|
377
|
+
"""
|
378
|
+
Get a specific BibTeX entry by its key
|
379
|
+
|
380
|
+
Args:
|
381
|
+
entry_key: The BibTeX entry key (e.g., "ACISP:LZXSW24")
|
382
|
+
force_download: Force download the newest crypto.bib file
|
383
|
+
|
384
|
+
Returns:
|
385
|
+
str: The BibTeX entry or None if not found
|
386
|
+
"""
|
387
|
+
try:
|
388
|
+
# Ensure we have the bib file locally
|
389
|
+
if not self._download_bib_file(force_download=force_download):
|
390
|
+
logger.error("Failed to download crypto.bib file")
|
391
|
+
return None
|
392
|
+
|
393
|
+
logger.info(f"Searching for BibTeX entry: {entry_key}")
|
394
|
+
current_entry = ""
|
395
|
+
in_entry = False
|
396
|
+
brace_count = 0
|
397
|
+
|
398
|
+
with open(self.bib_file_path, "r", encoding="utf-8") as f:
|
399
|
+
for line in f:
|
400
|
+
# Check if this is the start of the entry we're looking for
|
401
|
+
if line.strip().startswith(f"@") and entry_key in line:
|
402
|
+
current_entry = line
|
403
|
+
in_entry = True
|
404
|
+
brace_count = line.count("{") - line.count("}")
|
405
|
+
elif in_entry:
|
406
|
+
current_entry += line
|
407
|
+
brace_count += line.count("{") - line.count("}")
|
408
|
+
|
409
|
+
# Check if entry is complete
|
410
|
+
if brace_count <= 0:
|
411
|
+
return current_entry.strip()
|
412
|
+
|
413
|
+
except FileNotFoundError:
|
414
|
+
logger.error(f"crypto.bib file not found at {self.bib_file_path}")
|
415
|
+
except Exception as e:
|
416
|
+
logger.error(f"Error searching for entry {entry_key}: {e}")
|
417
|
+
|
418
|
+
return None
|
419
|
+
|
420
|
+
def _entry_matches_year_range(
|
421
|
+
self, bibtex_entry: str, year_min: int | None, year_max: int | None
|
422
|
+
) -> bool:
|
423
|
+
"""
|
424
|
+
Check if a BibTeX entry falls within the specified year range
|
425
|
+
|
426
|
+
Args:
|
427
|
+
bibtex_entry: Raw BibTeX entry text
|
428
|
+
year_min: Minimum year (inclusive, None means no minimum)
|
429
|
+
year_max: Maximum year (inclusive, None means no maximum)
|
430
|
+
|
431
|
+
Returns:
|
432
|
+
bool: True if entry is within year range, False otherwise
|
433
|
+
"""
|
434
|
+
# If no year constraints specified, all entries match
|
435
|
+
if year_min is None and year_max is None:
|
436
|
+
return True
|
437
|
+
|
438
|
+
try:
|
439
|
+
# Extract year from the BibTeX entry
|
440
|
+
year_match = re.search(
|
441
|
+
r'year\s*=\s*(?:["{\s]*)?(\d{4})', bibtex_entry, re.IGNORECASE
|
442
|
+
)
|
443
|
+
if not year_match:
|
444
|
+
# If no year found, exclude from results when year filtering is requested
|
445
|
+
return False
|
446
|
+
|
447
|
+
entry_year = int(year_match.group(1))
|
448
|
+
|
449
|
+
# Check minimum year constraint
|
450
|
+
if year_min is not None and entry_year < year_min:
|
451
|
+
return False
|
452
|
+
|
453
|
+
# Check maximum year constraint
|
454
|
+
if year_max is not None and entry_year > year_max:
|
455
|
+
return False
|
456
|
+
|
457
|
+
return True
|
458
|
+
|
459
|
+
except (ValueError, AttributeError):
|
460
|
+
# If year parsing fails, exclude from results when year filtering is requested
|
461
|
+
return False
|
@@ -5,13 +5,15 @@ import mcp.types as types
|
|
5
5
|
from mcp.server import NotificationOptions, Server
|
6
6
|
from mcp.server.models import InitializationOptions
|
7
7
|
|
8
|
-
# Import
|
8
|
+
# Import searchers
|
9
9
|
from .academic_platforms.iacr import IACRSearcher
|
10
|
+
from .academic_platforms.cryptobib import CryptoBibSearcher
|
10
11
|
|
11
12
|
server = Server("all-in-mcp")
|
12
13
|
|
13
|
-
# Initialize
|
14
|
+
# Initialize searchers
|
14
15
|
iacr_searcher = IACRSearcher()
|
16
|
+
cryptobib_searcher = CryptoBibSearcher(cache_dir="./downloads")
|
15
17
|
|
16
18
|
|
17
19
|
@server.list_tools()
|
@@ -83,6 +85,43 @@ async def handle_list_tools() -> list[types.Tool]:
|
|
83
85
|
"required": ["paper_id"],
|
84
86
|
},
|
85
87
|
),
|
88
|
+
types.Tool(
|
89
|
+
name="search-cryptobib-papers",
|
90
|
+
description="Search CryptoBib bibliography database for cryptography papers",
|
91
|
+
inputSchema={
|
92
|
+
"type": "object",
|
93
|
+
"properties": {
|
94
|
+
"query": {
|
95
|
+
"type": "string",
|
96
|
+
"description": "Search query string (e.g., 'cryptography', 'lattice', 'homomorphic')",
|
97
|
+
},
|
98
|
+
"max_results": {
|
99
|
+
"type": "integer",
|
100
|
+
"description": "Maximum number of papers to return (default: 10)",
|
101
|
+
"default": 10,
|
102
|
+
},
|
103
|
+
"return_bibtex": {
|
104
|
+
"type": "boolean",
|
105
|
+
"description": "Whether to return raw BibTeX entries (default: False)",
|
106
|
+
"default": False,
|
107
|
+
},
|
108
|
+
"force_download": {
|
109
|
+
"type": "boolean",
|
110
|
+
"description": "Force download the newest crypto.bib file (default: False)",
|
111
|
+
"default": False,
|
112
|
+
},
|
113
|
+
"year_min": {
|
114
|
+
"type": "integer",
|
115
|
+
"description": "Minimum publication year (inclusive, optional)",
|
116
|
+
},
|
117
|
+
"year_max": {
|
118
|
+
"type": "integer",
|
119
|
+
"description": "Maximum publication year (inclusive, optional)",
|
120
|
+
},
|
121
|
+
},
|
122
|
+
"required": ["query"],
|
123
|
+
},
|
124
|
+
),
|
86
125
|
]
|
87
126
|
|
88
127
|
|
@@ -186,6 +225,104 @@ async def handle_call_tool(
|
|
186
225
|
else:
|
187
226
|
return [types.TextContent(type="text", text=result)]
|
188
227
|
|
228
|
+
elif name == "search-cryptobib-papers":
|
229
|
+
query = arguments.get("query", "")
|
230
|
+
max_results = arguments.get("max_results", 10)
|
231
|
+
return_bibtex = arguments.get("return_bibtex", False)
|
232
|
+
force_download = arguments.get("force_download", False)
|
233
|
+
year_min = arguments.get("year_min")
|
234
|
+
year_max = arguments.get("year_max")
|
235
|
+
|
236
|
+
if not query:
|
237
|
+
return [
|
238
|
+
types.TextContent(
|
239
|
+
type="text", text="Error: Query parameter is required"
|
240
|
+
)
|
241
|
+
]
|
242
|
+
|
243
|
+
if return_bibtex:
|
244
|
+
# Return raw BibTeX entries
|
245
|
+
bibtex_entries = cryptobib_searcher.search_bibtex(
|
246
|
+
query,
|
247
|
+
max_results,
|
248
|
+
force_download=force_download,
|
249
|
+
year_min=year_min,
|
250
|
+
year_max=year_max,
|
251
|
+
)
|
252
|
+
|
253
|
+
if not bibtex_entries:
|
254
|
+
year_filter_msg = ""
|
255
|
+
if year_min or year_max:
|
256
|
+
year_range = (
|
257
|
+
f" ({year_min or 'earliest'}-{year_max or 'latest'})"
|
258
|
+
)
|
259
|
+
year_filter_msg = f" in year range{year_range}"
|
260
|
+
return [
|
261
|
+
types.TextContent(
|
262
|
+
type="text",
|
263
|
+
text=f"No BibTeX entries found for query: {query}{year_filter_msg}",
|
264
|
+
)
|
265
|
+
]
|
266
|
+
|
267
|
+
year_filter_msg = ""
|
268
|
+
if year_min or year_max:
|
269
|
+
year_range = f" ({year_min or 'earliest'}-{year_max or 'latest'})"
|
270
|
+
year_filter_msg = f" in year range{year_range}"
|
271
|
+
result_text = f"Found {len(bibtex_entries)} BibTeX entries for query '{query}'{year_filter_msg}:\n\n"
|
272
|
+
for i, entry in enumerate(bibtex_entries, 1):
|
273
|
+
result_text += f"Entry {i}:\n```bibtex\n{entry}\n```\n\n"
|
274
|
+
|
275
|
+
return [types.TextContent(type="text", text=result_text)]
|
276
|
+
else:
|
277
|
+
# Return parsed Paper objects
|
278
|
+
papers = cryptobib_searcher.search(
|
279
|
+
query,
|
280
|
+
max_results,
|
281
|
+
force_download=force_download,
|
282
|
+
year_min=year_min,
|
283
|
+
year_max=year_max,
|
284
|
+
)
|
285
|
+
|
286
|
+
if not papers:
|
287
|
+
year_filter_msg = ""
|
288
|
+
if year_min or year_max:
|
289
|
+
year_range = (
|
290
|
+
f" ({year_min or 'earliest'}-{year_max or 'latest'})"
|
291
|
+
)
|
292
|
+
year_filter_msg = f" in year range{year_range}"
|
293
|
+
return [
|
294
|
+
types.TextContent(
|
295
|
+
type="text",
|
296
|
+
text=f"No papers found for query: {query}{year_filter_msg}",
|
297
|
+
)
|
298
|
+
]
|
299
|
+
|
300
|
+
year_filter_msg = ""
|
301
|
+
if year_min or year_max:
|
302
|
+
year_range = f" ({year_min or 'earliest'}-{year_max or 'latest'})"
|
303
|
+
year_filter_msg = f" in year range{year_range}"
|
304
|
+
result_text = f"Found {len(papers)} CryptoBib papers for query '{query}'{year_filter_msg}:\n\n"
|
305
|
+
for i, paper in enumerate(papers, 1):
|
306
|
+
result_text += f"{i}. **{paper.title}**\n"
|
307
|
+
result_text += f" - Entry Key: {paper.paper_id}\n"
|
308
|
+
result_text += f" - Authors: {', '.join(paper.authors)}\n"
|
309
|
+
if paper.extra and "venue" in paper.extra:
|
310
|
+
result_text += f" - Venue: {paper.extra['venue']}\n"
|
311
|
+
if paper.published_date and paper.published_date.year > 1900:
|
312
|
+
result_text += f" - Year: {paper.published_date.year}\n"
|
313
|
+
if paper.doi:
|
314
|
+
result_text += f" - DOI: {paper.doi}\n"
|
315
|
+
if paper.extra and "pages" in paper.extra:
|
316
|
+
result_text += f" - Pages: {paper.extra['pages']}\n"
|
317
|
+
# Only include BibTeX when explicitly requested
|
318
|
+
if return_bibtex and paper.extra and "bibtex" in paper.extra:
|
319
|
+
result_text += (
|
320
|
+
f" - BibTeX:\n```bibtex\n{paper.extra['bibtex']}\n```\n"
|
321
|
+
)
|
322
|
+
result_text += "\n"
|
323
|
+
|
324
|
+
return [types.TextContent(type="text", text=result_text)]
|
325
|
+
|
189
326
|
else:
|
190
327
|
raise ValueError(f"Unknown tool: {name}")
|
191
328
|
|
@@ -1 +0,0 @@
|
|
1
|
-
# all_in_mcp/academic_platforms/__init__.py
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|