写了一个获取最新 nupkg url 的函数:
(defun lsp-python-ms-latest-nupkg-url (&optional channel)
(let ((channel (or channel "stable")))
(unless (member channel '("stable" "beta" "daily"))
(error (format "Unknown channel: %s" cnannel)))
(with-temp-buffer
(url-insert-file-contents
(concat
"https://pvsc.blob.core.windows.net/python-language-server-"
channel
"?restype=container&comp=list&prefix=Python-Language-Server-"
(cond (*macos* "osx")
(*linux* "linux")
(*winnt* "win")
(t (error (format "Unknown system: %s" system-type))))
"-x64"))
(pcase (xml-parse-region (point) (point-max))
(`((EnumerationResults
((ContainerName . ,_))
(Prefix nil ,_)
(Blobs nil . ,blobs)
(NextMarker nil)))
(cdar
(sort
(mapcar (lambda (blob)
(pcase blob
(`(Blob
nil
(Name nil ,_)
(Url nil ,url)
(Properties nil (Last-Modified nil ,last-modified) . ,_))
(cons (encode-time (parse-time-string last-modified)) url))))
blobs)
(lambda (t1 t2)
(time-less-p (car t2) (car t1))))))))))
把中间判断 system-type 那几行替换城你自己的应该就可以了,以下是在 macOS 的测试结果:
(list
(lsp-python-ms-latest-nupkg-url)
(lsp-python-ms-latest-nupkg-url "beta")
(lsp-python-ms-latest-nupkg-url "daily"))
;; =>
;; ("https://pvsc.blob.core.windows.net/python-language-server-stable/Python-Language-Server-osx-x64.0.3.20.nupkg"
;; "https://pvsc.blob.core.windows.net/python-language-server-beta/Python-Language-Server-osx-x64.0.3.20.nupkg"
;; "https://pvsc.blob.core.windows.net/python-language-server-daily/Python-Language-Server-osx-x64.0.3.20.nupkg")
另外需要特别注意的是,输出的原始 xml 里有个 NextMarker 好像是分页标记,我目前直接当作 nil 处理没有问题(如果为 t 则 pcase pattern 不起作用),说明目前是没有分页的。将来有可能会返回分页数据,不过也可能源头控制好了,不产生分页数据,否则 daily 每天一个包早就该分页了。
UPDATE: 2019-06-27 01.43.47
修复排序问题 154楼
UPDATE: 2019-06-28 15.54.21
Replace url-retrieve-synchronously with url-insert-file-contents