aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/importers/cdl_dash_dat.py
blob: c1ea075dc44c0fec7ffc9d15d48ea8f0995f164a (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
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
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
#!/usr/bin/env python3

import os
import sys
import json
import magic
import urllib
import hashlib
import mimetypes
import subprocess

import fatcat_client
from fatcat_client import *
from .common import clean
from .crossref import lookup_license_slug


def single_file(prefix, path):

    full = prefix + path
    size_bytes = os.stat(full).st_size

    hashes = [
        hashlib.md5(),
        hashlib.sha1(),
        hashlib.sha256(),
    ]
    with open(full, 'rb') as fp:
        while True:
            data = fp.read(2**20)
            if not data:
                break
            for h in hashes:
                h.update(data)
    mime = magic.Magic(mime=True).from_file(full)
    if mime == 'application/octet-stream':
        # magic apparently isn't that great; try using filename as well
        guess = mimetypes.guess_type(full)[0]
        if guess:
            mime = guess

    fsm = FilesetEntityManifest(
        path=path,
        size=size_bytes,
        md5=hashes[0].hexdigest(),
        sha1=hashes[1].hexdigest(),
        sha256=hashes[2].hexdigest(),
        extra=dict(mimetype=mime))
    return fsm

def make_manifest(base_dir):
    manifest = []
    for root, dirs, files in os.walk(base_dir):
        for f in files:
            manifest.append(single_file(root, f))
    return manifest


def cdl_dash_release(meta, extra=None):

    if not extra:
        extra = dict()

    assert meta['identifier']['type'] == 'DOI'
    doi = meta['identifier']['value'].lower()
    assert doi.startswith('10.')

    ark_id = None
    for extid in meta.get('alternativeIdentifiers', []):
        if extid['value'].startswith('ark:'):
            ark_id = extid['value']
    assert ark_id

    license_slug = lookup_license_slug(meta['rights']['uri'])

    abstracts = []
    for desc in meta['descriptions']:
        if desc['type'] == "abstract":
            abstracts.append(ReleaseEntityAbstracts(
                mimetype="text/html",
                content=clean(desc['value'])))
            #print(abstracts)
    if not abstracts:
        abstracts = None
    
    contribs = []
    for creator in meta['creator']:
        contribs.append(ReleaseContrib(
            # TODO: given_name=creator['given'],
            # TODO: surname=creator['family'],
            # sorry everybody
            raw_name="{} {}".format(creator['given'], creator['family']),
            raw_affiliation=creator.get('affiliation'),
            role="author", # presumably, for these datasets?
        ))

    r = ReleaseEntity(
        ext_ids=ReleaseEntityExtIds(
            doi=doi,
            ark=ark_id,
        ),
        title=clean(meta['title'], force_xml=True),
        publisher=clean(meta['publisher']),
        release_year=int(meta['publicationYear']),
        release_type="dataset",
        license_slug=license_slug,
        contribs=contribs,
        abstracts=abstracts,
        extra=extra,
    )
    return r

def make_release_fileset(dat_path):

    if dat_path.endswith('/'):
        dat_path = dat_path[:-1]
    dat_discovery = dat_path
    extra = dict()
    assert len(dat_discovery) == 64

    with open(dat_path + "/cdl_dash_metadata.json", 'r') as fp:
        meta_dict = json.loads(fp.read())
    
    release = cdl_dash_release(meta_dict)
    ark_id = release.extra['ark_id']

    dash_version = None
    # really crude XML parse-out
    with open(dat_path + "/stash-wrapper.xml", 'r') as fp:
        for line in fp:
            line = line.strip()
            if line.startswith("<st:version_number>"):
                dash_version = int(line[19:].split('<')[0])
    assert dash_version is not None
    extra['cdl_dash'] = dict(version=dash_version)
    release.extra['cdl_dash'] = dict(version=dash_version)

    manifest = make_manifest(dat_path + "/files/")

    bundle_url = dict(
        url="https://merritt.cdlib.org/u/{}/{}".format(
            urllib.parse.quote(ark_id, safe=''),
            dash_version),
        rel="repo-bundle")
    repo_url = dict(
        url="https://merritt.cdlib.org/d/{}/{}/".format(
            urllib.parse.quote(ark_id, safe=''),
            dash_version),
        rel="repo")
    dat_url = dict(
        url="dat://{}/files/".format(dat_discovery),
        rel="dweb")
    fs = FilesetEntity(
        urls=[bundle_url, repo_url, dat_url],
        release_ids=None,
        manifest=manifest,
        extra=extra)
    return (release, fs)

def auto_cdl_dash_dat(api, dat_path, release_id=None, editgroup_id=None):

    git_rev = subprocess.check_output(
        ["git", "describe", "--always"]).strip().decode('utf-8')

    (release, fileset) = make_release_fileset(dat_path)

    if not editgroup_id:
        eg = api.create_editgroup(Editgroup(
            description="One-off import of dataset(s) from CDL/DASH repository (via IA, Dat dweb pilot project)",
            extra=dict(
                git_rev=git_rev,
                agent="fatcat_tools.auto_cdl_dash_dat")))
        editgroup_id = eg.editgroup_id

    if not release_id and release.ext_ids.doi:
        try:
            r = api.lookup_release(doi=release.ext_ids.doi)
            release_id = r.ident
        except fatcat_client.rest.ApiException:
            pass
    if not release_id:
        edit = api.create_release(release, editgroup_id=editgroup_id)
        release_id = edit.ident

    release = api.get_release(release_id, expand="filesets")
    if len(release.filesets):
        print("A fileset already exists for release {}".format(release.ident))
        return (None, None, None)

    fileset.release_ids = [release.ident]
    edit = api.create_fileset(fileset, editgroup_id=editgroup_id)
    fileset = api.get_fileset(edit.ident)
    return (editgroup_id, release, fileset)

if __name__=='__main__':
    # pass this a discovery key that has been cloned to the local directory
    print(make_release_fileset(sys.argv[1]))