* contrib/lisp/org-element.el (org-element-recursive-objects): Add
macro object type to recursive types.
(org-element-object-restrictions): Macros can only contain other
macros.
* contrib/lisp/org-export.el (org-export-get-inbuffer-options): Parse
macro value, unless it it meant to be eval'ed.
(org-export-expand-macro): Expand macros recursively.
* contrib/lisp/org-element.el (org-element-link-parser): Do not parse
radio link as a recursive link.
* contrib/lisp/org-export.el (org-export-solidify-link-text): Make
sure "\" isn't allowed in solidified strings. This is required to
allow entities in radio targets.
* EXPERIMENTAL/org-e-latex.el (org-e-latex-link): Apply changes to
radio links.
* EXPERIMENTAL/org-e-ascii.el (org-e-ascii-link): Apply changes to
radio links.
* contrib/lisp/org-element.el (org-element-object-restrictions): Allow
plain links in link description.
* contrib/lisp/org-export.el (org-export-default-inline-image-rule):
New variable.
(org-export-inline-image-p): Use rules instead of extensions. A rule
is a regexp applied to path for a given type of link. It allows to
extend inline images to non-local files.
* EXPERIMENTAL/org-e-latex.el (org-e-latex-inline-image-rules): New
variable.
(org-e-latex-inline-image-extensions): Removed variable
There are two motivations behind this change. The first one is to
allow, for example, an HTML exporter to define a rule like the
following:
("http" . "\\.\\(png\\|jpg\\|jpeg\\|gif\\)\\(\\?\\(\\w+=\\w+&?\\)*\\)?")
The other one is to properly define thumbnails (clickable images),
like the following:
[[http://orgmode.org][file:~/my-logo.png]]
* contrib/lisp/org-element.el (org-element-map): Remove included file
expansion part.
* contrib/lisp/org-export.el (org-export-as): Expand include keywords
before executing blocks.
(org-export-expand-include-keyword, org-export-prepare-file-contents):
New functions.
(org-export-included-file, org-export-parse-included-file): Removed
functions.
* EXPERIMENTAL/org-e-ascii.el (org-e-ascii-keyword): Remove include
keyword handling.
* EXPERIMENTAL/org-e-latex.el (org-e-latex-keyword): Remove include
keyword handling.
Back-ends do not need anymore to take care of #+include keywords.
This change is required since file inclusion can potentially break any
structure. Hence, it should be done before parsing.
* contrib/lisp/org-element.el (org-element-item-parser): If item's
first line is empty, ensure contents start at a beginning of line so
that child can be correctly parsed.
* contrib/lisp/org-element.el (org-element-dual-keywords): Move
caption in the dual keywords category.
(org-element--affiliated-re): Change regexp matching a dual keyword.
(org-element-collect-affiliated-keywords): Dual parsed keywords also
get their optional value parsed as a secondary string.
* EXPERIMENTAL/org-e-latex.el (org-e-latex--caption/label-string,
org-e-latex-src-block): Apply caption status change.
* EXPERIMENTAL/org-e-ascii.el (org-e-ascii--build-caption): Apply
caption status change.
(org-e-ascii--list-listings, org-e-ascii--list-tables): Apply caption
status change. Also use short caption name when available.
The status change implies a syntax change. Captions can now share
#+results: syntax, which mean that they can have a secondary
value. Thus, the following Org snippets are valid, the first two being
equivalent.
#+caption: long name
#+caption[]: long name
#+caption[short name]: long name
* contrib/lisp/org-element.el (org-element-quote-section-parser):
Assume point is at quote-section beginning.
(org-element-guess-type): Move point to section beginning before
calling for quote-section parser.
(org-element-parse-elements, org-element-current-element): Correctly
detect and parse quote section.
* contrib/lisp/org-element.el (org-element--element-block-re): New
variable.
(org-element-current-element): New function.
(org-element-parse-elements): Make use of the new specialized function
instead of `org-element-at-point'. Also narrow buffer to current
container (greater element or recursive object) being parsed to
improve search speed.
(org-element-comment-parser): Speed-up parsing for comments at column 0.
(org-element-guess-type): Make comment regexp less restrictive, due to
comment optimizations.
* contrib/lisp/org-element.el (org-element-parse-objects): Untabify
strings between objects to avoid any `tab-width' mismatch.
* contrib/lisp/org-element.el (org-element-normalize-contents): No
longer need to handle tabs.
* contrib/lisp/org-element.el (org-element-map): Remove use of
`org-export-update-info'.
* contrib/lisp/org-export.el (org-export-persistent-properties,
org-export-persistent-properties-list): Remove variables
(org-export-collect-tree-properties): Rename from
`org-export-initialize-persistent-properties'.
(org-export-data): Get rid of `org-export-update-info' calls.
(org-export-as): Use new `org-export-collect-tree-properties' name.
(org-export-resolve-coderef, org-export-get-loc): New functions.
(org-export-handle-code): Use new functions instead of removed
properties. Reformat code. Change signature.
* contrib/lisp/org-element.el (org-element-section-parser,
org-element-section-interpreter): New functions
(org-element-greater-elements): Add new element to the list.
(org-element-at-point): Change arguments to handle a new section mode.
(org-element-guess-type): Accept an optional argument to look for
sections in priority.
(org-element-parse-buffer): Start in section mode by default. Thus
any text before the first headline is still in a section of his
own.
(org-element-parse-elements): Handle new section mode.
* contrib/lisp/org-element.el (org-element-headline-parser): Update
list of attributes in the docstring.
(org-element-inlinetask-parser): Complete attributes with properties
inherited from any property drawer within the inlinetask and with
time properties (i.e. scheduled, clock...).
* contrib/lisp/org-element.el (org-element-map): Change name of local
variables to prevent variable name collisions with the function
around org-element-map.
Nicolas Goaziou
Bastien Guerry