JSONFeeds, JSON scraping, and POST requests for feeds (#5662)

* allow POST requests for feeds

* added json dotpath and jsonfeed subscriptions. No translation strings yet

* debug and fix jsonfeed parser

* bugfix params saved when editing feed

* added translations for JSON features

* Update docs for web scraping

* make fix-all
and revert unrelated changes, plus a few manual fixes, but there are still several type errors

* Fix some i18n

* refactor json parsing for both feed types

* cleanup unnecessary comment

* refactored generation of SimplePie for XPath and JSON feeds

* Fix merge error

* Update to newer FreshRSS code

* A bit of refactoring

* doc, whitespace

* JSON Feed is in two words

* Add support for array syntax

* Whitespace

* Add OPML export/import

* Work on i18n

* Accept application/feed+json

* Rework POST

* Fix update

* OPML for cURL options

* Fix types

* Fix Typos

---------

Co-authored-by: Erion Elmasllari <elmasllari@factorsixty.com>
Co-authored-by: Alexandre Alapetite <alexandre@alapetite.fr>

2 files changed, 76 insertions, 3 deletions

diff --git a/docs/en/developers/OPML.md b/docs/en/developers/OPML.md
index 5191592a8..0cc7cabbf 100644
--- a/docs/en/developers/OPML.md
+++ b/docs/en/developers/OPML.md
@@ -44,6 +44,44 @@ The following attributes are using similar naming conventions than [RSS-Bridge](
 * `frss:xPathItemCategories`: XPath expression for extracting a list of categories (tags) from the item context.
 * `frss:xPathItemUid`: XPath expression for extracting an item’s unique ID from the item context. If left empty, a hash is computed automatically.
 
+### JSON+DotPath
+
+* `<outline type="JSON+DotPath" ...`: Similar to `HTML+XPath` but for JSON and using a dot/bracket syntax such as `object.object.array[2].property`.
+
+* `frss:jsonItem`: JSON dot path for extracting the feed items from the source page.
+	* Example: `data.items`
+* `frss:jsonItemTitle`: JSON dot path for extracting the item’s title from the item context.
+	* Example: `meta.title`
+* `frss:jsonItemContent`: JSON dot path for extracting an item’s content from the item context.
+	* Example: `content`
+* `frss:jsonItemUri`: JSON dot path for extracting an item link from the item context.
+	* Example: `meta.links[0]`
+* `frss:jsonItemAuthor`: JSON dot path for extracting an item author from the item context.
+* `frss:jsonItemTimestamp`: JSON dot path for extracting an item timestamp from the item context. The result will be parsed by [`strtotime()`](https://php.net/strtotime).
+* `frss:jsonItemTimeFormat`: Date/Time format to parse the timestamp, according to [`DateTime::createFromFormat()`](https://php.net/datetime.createfromformat).
+* `frss:jsonItemThumbnail`: JSON dot path for extracting an item’s thumbnail (image) URL from the item context.
+* `frss:jsonItemCategories`: JSON dot path for extracting a list of categories (tags) from the item context.
+* `frss:jsonItemUid`: JSON dot path for extracting an item’s unique ID from the item context. If left empty, a hash is computed automatically.
+
+### JSON Feed
+
+* `<outline type="JSONFeed" ...`: Uses `JSON+DotPath` behind the scenes to parse a [JSON Feed](https://www.jsonfeed.org/).
+
+### cURL
+
+A number of [cURL options](https://curl.se/libcurl/c/curl_easy_setopt.html) are supported:
+
+* `frss:CURLOPT_COOKIE`
+* `frss:CURLOPT_COOKIEFILE`
+* `frss:CURLOPT_FOLLOWLOCATION`
+* `frss:CURLOPT_HTTPHEADER`
+* `frss:CURLOPT_MAXREDIRS`
+* `frss:CURLOPT_POST`
+* `frss:CURLOPT_POSTFIELDS`
+* `frss:CURLOPT_PROXY`
+* `frss:CURLOPT_PROXYTYPE`
+* `frss:CURLOPT_USERAGENT`
+
 ### Miscellaneous
 
 * `frss:cssFullContent`: [CSS Selector](https://developer.mozilla.org/en-US/docs/Web/CSS/CSS_Selectors) to enable the download and extraction of the matching HTML section of each articles’ Web address.
diff --git a/docs/en/users/11_website_scraping.md b/docs/en/users/11_website_scraping.md
index f51c2ba40..f725280ea 100644
--- a/docs/en/users/11_website_scraping.md
+++ b/docs/en/users/11_website_scraping.md
@@ -5,9 +5,14 @@ FreshRSS has a built-in [Web scraping](https://en.wikipedia.org/wiki/Web_scrapin
 ## How to add
 
 Go to “Subscription Management” where a new feed can be added.
-Change the “Type of feed source” to “HTML + XPath (Web scraping)”.
-An additional list of text boxes to configure the web scraping.
-[XPath 1.0](https://www.w3.org/TR/xpath-10/) is used as traversing language.
+Change the “Type of feed source” to one of:
+- “HTML + XPath (Web scraping)”
+- JSON Feed (see [`jsonfeed.org`](https://www.jsonfeed.org/))
+- JSON (Dotted paths)
+
+An additional list of text boxes to configure the Web scraping will show.
+
+For HTML + XPath, [XPath 1.0](https://www.w3.org/TR/xpath-10/) is used as traversing language.
 
 ### Get the XPath path
 
@@ -15,6 +20,36 @@ Firefox: the built-in “inspect” tool may be used to help create a valid XPat
 Select the node in the HTML, right click with your mouse and chose “Copy” and “XPath”.
 The XPath is stored in your clipboard now.
 
+### Get the JSON dotted path
+
+Suppose the JSON to which you are subscribing to (or scraping) looks like this:
+
+```json
+{
+	"data": {
+		"items": [
+			{
+				"meta": {"title": "Some news item"},
+				"content": "Content of the news",
+				"links": ["https://example.net/1", "https://example.org/1"]
+			},
+			{
+				"meta": {"title": "Some other news item"},
+				"content": "Yet more content",
+				"links": ["https://example.net/2", "https://example.org/2"]
+			}
+		]
+	}
+}
+```
+
+The *dot notation* and *bracket notation* (only numeric) are supported.
+
+Then the items are under `data.items`, and within each item, the title is `meta.title`,
+and the link would be `links[1]`.
+
+It is a similar syntax to the JavaScript way to access JSON: `object.object.array[2].property`.
+
 ## Tips & tricks
 
 - [Timezone of date](https://github.com/FreshRSS/FreshRSS/discussions/5483)