forked from ribice/glice
-
Notifications
You must be signed in to change notification settings - Fork 0
/
repository.go
289 lines (250 loc) · 7.03 KB
/
repository.go
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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
package glice
import (
"context"
"fmt"
"net/http"
"net/url"
"strings"
)
const (
htmlRegion = `head`
htmlTag = `meta`
sourceName = "go-source"
importName = "go-import"
)
type Repositories = []*Repository
var _ RepoInfoGetter = (*Repository)(nil)
// Repository holds information about the repository
type Repository struct {
Import string
Imports []string
url *url.URL
license *RepositoryLicense
parsed bool
resolved bool
}
// ScanRepositories scans the directory providing in options and returns
// a slice of *Repository objects
func ScanRepositories(ctx context.Context, options *Options) ([]*Repository, error) {
var repos Repositories
var el ErrorList
modules, err := ParseModFile(
options.SourceDir,
!options.DirectOnly,
)
if err != nil {
err = fmt.Errorf("unable to list repositories for '%s'; %w",
options.SourceDir,
err)
goto end
}
repos = make(Repositories, len(modules))
el = make(ErrorList, 0)
for i, module := range modules {
r := &Repository{
Import: module,
}
Infof("Resolving repository for '%s'", r.Import)
err = r.ResolveRepository(ctx, options)
if err != nil {
el = append(el, err)
continue
}
Infof("\tas '%s'", r.GetURL())
repos[i] = r
}
if len(el) != 0 {
err = el
}
end:
return repos, err
}
// ResolveRepository accepts a Go import string and 1. parses it as a URL
// then either 2. classifies it at a known repo domain, or 3. uses an
// HTTPS request to resolve it from HTML <meta> tags in the response.
// See https://pkg.go.dev/cmd/go#hdr-Remote_import_paths
func (r *Repository) ResolveRepository(ctx context.Context, options *Options) (err error) {
// Capture the original import as it will get changed if we need
// recursion, but we want the original one on final return.
imp := r.Import
// Add an HTTPS scheme and use *url.ParseModFile() to validate to be
// a valid URL, If valid, assign returned *url.URL to r.url.
err = r.ParseImport(imp)
if err != nil {
err = fmt.Errorf("invalid repository import '%s'; %w", r.Import, err)
goto end
}
if r.KnownDomainRecognized() {
// It is a known repository domain (github.com, etc.) so no need
// to resolve an HTTP request to determine the repo URL.
// Note at this point in repo resolution this is about the host of
// the Import, not the host of the RepoURL
goto end
}
// HTTPS request the import as a URL then parse the HTML <meta> tags to determine
// its repo URL as per https://pkg.go.dev/cmd/go#hdr-Remote_import_paths
// See paragraph starting with "For code hosted on other servers"
err = r.ResolveImport(ctx, options)
if err != nil {
err = fmt.Errorf("unable to resolve repository import '%s'; %w", r.Import, err)
}
end:
// Restore the value of the original import
r.Import = imp
return err
}
// SetImport sets a new Import, but keeps track of prior Import values
func (r *Repository) SetImport(imp string) {
r.Import = imp
if r.Imports == nil {
r.Imports = []string{imp}
}
if imp != r.Imports[len(r.Imports)-1] {
r.Imports = append(r.Imports, imp)
}
}
func (r *Repository) GetOrgName() string {
return ExtractFieldWithDelimiter(r.GetPath(), 1, '/')
}
func (r *Repository) GetRepoName() string {
return ExtractFieldWithDelimiter(r.GetPath(), 2, '/')
}
func (r *Repository) GetRepoURL() string {
r.checkURL()
return r.url.String()
}
func (r *Repository) GetHost() string {
r.checkURL()
return r.url.Host
}
func (r *Repository) GetPath() string {
r.checkURL()
return strings.TrimLeft(r.url.Path, "/")
}
func (r *Repository) GetURL() string {
r.checkURL()
return r.url.String()
}
func (r *Repository) checkURL() {
if r.url == nil {
panic(fmt.Sprintf("Cannot call Repository.%s() before calling Repository.ResolveRepository()",
CallerName()))
}
}
func (r *Repository) GetLicenseID() string {
if r.license == nil {
return "License inaccessible"
}
return r.license.ID
}
func (r *Repository) GetLicenseURL() string {
if r.license == nil {
return "http://inaccessible"
}
return r.license.URL
}
// ParseImport parses an import string which is typically a URL w/o the scheme So this func
// adds a scheme and the parses the URL using *url.Parse() from the "net/url" import.
func (r *Repository) ParseImport(imp string) (err error) {
r.SetImport(imp)
locator := fmt.Sprintf("https://%s", r.Import)
err = r.ParseURL(locator)
if err != nil {
err = fmt.Errorf("invalid repository import: '%s'; %w", r.Import, err)
}
r.parsed = true
return err
}
// ParseURL parses a URL in string format into a *url.URL and assigns
// to Repository.url if successful, or returns an error if not
func (r *Repository) ParseURL(u string) (err error) {
r.url, err = url.Parse(u)
if err != nil {
err = fmt.Errorf("invalid repository URL: '%s'; %w", u, err)
r.url = &url.URL{}
}
return err
}
var importCache = map[string]*url.URL{}
// ResolveImport indirect repos as described here:
// https://golang.org/cmd/go/#hdr-Remote_import_paths
func (r *Repository) ResolveImport(ctx context.Context, options *Options) (err error) {
var resp *http.Response
var meta *Meta
var repoURL string
var imp = r.Import
if repoURL, ok := importCache[imp]; ok {
r.url = repoURL
goto end
}
resp, err = HTTPGetWithContext(ctx, fmt.Sprintf("%s?go-get=1", r.GetURL()))
if err != nil {
err = fmt.Errorf("failed to retrieve %s; %w", r.GetURL(), err)
goto end
}
defer MustClose(resp.Body)
meta, err = GetMetaFromHTMLReader(resp.Body)
if err != nil {
err = fmt.Errorf(`failed to extract <meta> from HTML at %s; %w`,
r.GetURL(),
err)
goto end
}
repoURL, err = meta.ResolveGoImport(r.Import)
if err == nil {
err = r.ParseURL(repoURL)
if err != nil {
err = fmt.Errorf(`failed to parse URL '%s' from <meta> in HTML at %s; %w`,
repoURL,
r.GetURL(),
err)
goto end
}
}
if err == ErrRequestPrefixInstead || err == nil {
r.SetImport(StripURLScheme(repoURL))
err = r.ResolveRepository(ctx, options)
if err != nil {
err = fmt.Errorf("unable to resolve repository '%s'; %w", repoURL, err)
goto end
}
}
if err != nil {
err = fmt.Errorf(`failed to parse <meta> from HTML at %s; %w`,
r.GetURL(),
err)
goto end
}
importCache[imp] = r.url
end:
return err
}
// ResolveLicense requests the license for the repository
func (r *Repository) ResolveLicense(ctx context.Context, options *Options) (err error) {
var ra RepositoryAdapter
if r.license != nil {
goto end
}
ra, err = GetRepositoryAdapter(ctx, r)
if err != nil {
Failf(ExitCannotGetRepositoryAdapter,
"unable to get repository getter for %s",
r.GetHost())
}
r.license, err = ra.GetRepositoryLicense(ctx, options)
if err != nil {
err = fmt.Errorf("unable to get license for '%s'; %w",
ra.GetName(),
err)
goto end
}
end:
return err
}
// KnownDomainRecognized inspects r.url.Host and returns true if the domain is
// recognized explicitly by this program, or false otherwise. If it is recognized
// and there are no <meta> tags for ?go-get=1 then it may also update the Host
// and Path of r.url, as applicable.
func (r *Repository) KnownDomainRecognized() (recognized bool) {
return GetRepositoryAdapterFunc(r) != nil
}