Skip to content

download_dblp

src.utils.apis.download_dblp

Download DBLP XML database (~3 GB compressed) for author matching.

Usage::

python -m src.utils.download_dblp            # interactive
python -m src.utils.download_dblp --auto     # non-interactive (CI)

Output: data/dblp/dblp.xml.gz

This replaces the former scripts/download_dblp.sh and removes the dependency on curl.

download_dblp(*, auto: bool = False) -> bool

Download or update the DBLP XML file.

Parameters

auto : bool If True run non-interactively (skip prompts, download only when missing or outdated).

Returns

bool True if the file is present and valid at exit.

Source code in src/utils/apis/download_dblp.py
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
def download_dblp(*, auto: bool = False) -> bool:
    """Download or update the DBLP XML file.

    Parameters
    ----------
    auto : bool
        If ``True`` run non-interactively (skip prompts, download only
        when missing or outdated).

    Returns
    -------
    bool
        ``True`` if the file is present and valid at exit.
    """
    DBLP_DIR.mkdir(parents=True, exist_ok=True)

    if DBLP_FILE.is_file():
        size_mb = DBLP_FILE.stat().st_size >> 20
        freshness = _is_up_to_date(DBLP_FILE)

        if freshness is True:
            logger.info("DBLP file is up to date (%d MB)", size_mb)
            return True

        if freshness is False:
            logger.warning("DBLP file is outdated (%d MB)", size_mb)
        else:
            logger.warning("DBLP file exists (%d MB), could not check remote date", size_mb)
            if auto:
                return True

        if not auto:
            try:
                answer = input("Re-download? (y/N): ").strip().lower()
            except EOFError:
                answer = "n"
            if answer != "y":
                return True

        DBLP_FILE.unlink()

    # Connectivity check
    try:
        headers = {"User-Agent": "ReproDB-Pipeline/1.0 (+https://github.com/ReproDB/reprodb-pipeline)"}
        requests.head(DBLP_URL, timeout=10, headers=headers)
    except requests.ConnectionError:
        proxy = os.environ.get("https_proxy", "")
        logger.error("Cannot connect to dblp.org (proxy: %s)", proxy)
        return False

    _download(DBLP_FILE)

    size_mb = DBLP_FILE.stat().st_size >> 20
    if size_mb < MIN_SIZE_MB:
        logger.error("File too small (%d MB, expected >= %d MB) — download may be truncated", size_mb, MIN_SIZE_MB)
        return False

    logger.info("Download complete (%d MB)", size_mb)
    return True