How to Discover and Simulate Baidu Cloud’s Transfer API for Automated File Saving

First, log into a Baidu Cloud account and open a shared link in a browser (Firefox is used in the example). Open the developer console (F12) and capture network traffic while manually performing a transfer: select all files → Save to cloud → Choose a destination folder → Confirm. Clear previous logs before clicking Confirm to isolate the transfer request.

The captured request is a POST containing the word “transfer” in its URL, which is the API we need. Analyzing the request headers and parameters is essential for code simulation.

Click the request and view its Cookies to see the required authentication cookies. Since the transfer operation requires a logged‑in state, the relevant cookies must be included in the simulated request.

Cookie analysis

Delete all Baidu‑related cookies in the browser (Settings → Privacy → Remove cookies). Then log in again and observe the remaining cookies. The BDUSS cookie disappears after deletion, indicating it carries the login session. Similarly, the STOKEN cookie is required; without it the transfer prompts for re‑login.

With the necessary cookies identified, construct the request headers accordingly.

Other header fields can be copied from the captured request; the two cookies must be kept up‑to‑date because they expire.

Parameter analysis

Inspect the request’s Params tab. The query string contains shareid, from, and bdstoken which must be supplied manually; the rest can be left unchanged. In the form data, filelist and path represent the source directory in the sharer’s cloud and the target directory you selected.

Thus the required parameters are: shareid, from, bdstoken, filelist, and path. bdstoken can be obtained from a manual transfer capture; path is user‑defined, assuming it exists in your cloud storage. The other three values are extracted from the share link.

Construct the transfer request URL using these parameters.

To obtain shareid, from, and filelist, fetch the shared page source and search for “shareid”. The relevant values appear in a JavaScript block; yunData.PATH gives the path, while yunData.FILEINFO contains a JSON‑encoded filelist with Unicode‑encoded filenames.

Because direct requests calls often return 404 without proper cookies, the author used Selenium’s WebDriver to perform two GET requests: the first obtains a BAIDUID cookie, and the second succeeds with the full cookie set.

The structure of yunData.FILEINFO can be inspected by pasting it into a JSON formatter.

Using regular expressions, extract the three parameters from the page source. The extracted values can then be passed to the previously built transfer method to automate the file saving.