Require a space after colon instead of tab for fixed width area.

docker/organic_test/Dockerfile

@@ -88,7 +88,7 @@ ARG DOOMEMACS_PATH=/foreign_documents/doomemacs
 ARG DOOMEMACS_REPO=https://github.com/doomemacs/doomemacs.git
 RUN mkdir -p $DOOMEMACS_PATH && git -C $DOOMEMACS_PATH init --initial-branch=main && git -C $DOOMEMACS_PATH remote add origin $DOOMEMACS_REPO && git -C $DOOMEMACS_PATH fetch origin $DOOMEMACS_VERSION && git -C $DOOMEMACS_PATH checkout FETCH_HEAD
 
-ARG WORG_VERSION=74e80b0f7600801b1d1594542602394c085cc2f9
+ARG WORG_VERSION=0c8d5679b536af450b61812246a3e02b8103f4b8
 ARG WORG_PATH=/foreign_documents/worg
 ARG WORG_REPO=https://git.sr.ht/~bzg/worg
 RUN mkdir -p $WORG_PATH && git -C $WORG_PATH init --initial-branch=main && git -C $WORG_PATH remote add origin $WORG_REPO && git -C $WORG_PATH fetch origin $WORG_VERSION && git -C $WORG_PATH checkout FETCH_HEAD

org_mode_samples/lesser_element/fixed_width_area/tab_instead_of_space.org

@@ -0,0 +1,2 @@
+# Fixed width areas must begin with colon followed by a space, not a tab, so this is not a fixed width area.
+:	foo

src/parser/fixed_width_area.rs

@@ -3,7 +3,6 @@ use nom::bytes::complete::is_not;
 use nom::bytes::complete::tag;
 use nom::character::complete::line_ending;
 use nom::character::complete::space0;
-use nom::character::complete::space1;
 use nom::combinator::eof;
 use nom::combinator::not;
 use nom::combinator::recognize;
@@ -12,6 +11,7 @@ use nom::sequence::preceded;
 use nom::sequence::tuple;
 
 use super::org_source::OrgSource;
+use super::util::only_space1;
 use super::util::org_line_ending;
 use crate::context::parser_with_context;
 use crate::context::RefContext;
@@ -50,7 +50,7 @@ fn fixed_width_area_line<'b, 'g, 'r, 's>(
     let (remaining, _indent) = space0(input)?;
     let (remaining, _) = tuple((
         tag(":"),
-        alt((recognize(tuple((space1, is_not("\r\n")))), space0)),
+        alt((recognize(tuple((only_space1, is_not("\r\n")))), space0)),
         org_line_ending,
     ))(remaining)?;
     let source = get_consumed(input, remaining);

src/parser/util.rs

@@ -1,4 +1,5 @@
 use nom::branch::alt;
+use nom::bytes::complete::is_a;
 use nom::character::complete::anychar;
 use nom::character::complete::line_ending;
 use nom::character::complete::none_of;
@@ -228,9 +229,16 @@ where
     }
 }
 
+/// Match at least one space character.
+///
+/// This is similar to nom's space1 parser except space1 matches both spaces and tabs whereas this only matches spaces.
+pub(crate) fn only_space1<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
+    is_a(" ")(input)
+}
+
 /// Match single space or tab.
 ///
-/// In org-mode syntax, spaces and tabs are interchangeable.
+/// In org-mode syntax, spaces and tabs are often (but not always!) interchangeable.
 pub(crate) fn org_space<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, char> {
     one_of(" \t")(input)
 }