Module markup 

Plain text from publisher metadata strings that carry inline markup.

Crossref serves titles as the publisher deposited them, and several publishers deposit JATS / MathML inline elements pretty-printed onto their own lines (#609):

"Recent developments in the P\n                    <scp>y</scp>\n                    SCF program package"

Stripping the tags is not enough. The newline-plus-indent runs around an inline element were added by the pretty-printer, so they say nothing about whether the author wrote a space there: P<scp>y</scp>SCF is one word, in <i>γ</i>-InSe is two. plain_title decides each such run from its neighbours; the rules are pinned against real deposits in the tests below.

Whitespace that is not a pretty-printer run (no line break, or no tag beside it) collapses to one space, and $...$ is left alone: a hand-written TeX title is content, which doiget lint already treats as intentional.

Functions

has_inline_markup

Whether s carries a line break or an inline markup tag, i.e. whether plain_title would rewrite more than entities. doiget lint uses this to find entries pasted from an older doiget cite (#609).

plain_title

raw with inline markup reduced to its text content, XML entities decoded and whitespace collapsed. A string with no tag, entity or line break is returned unchanged.