mirror of https://github.com/mpastell/Weave.jl
fix code chunk line numbers
Shuhei Kadowaki
parent
11c720ad7f
commit
2a6ee38850
|
|
@ -26,7 +26,7 @@ function restore_chunk(chunk::CodeChunk, cached)
|
|||
|
||||
# Chunk types, don't match after loading. Fix by constructing chunks
|
||||
# from loaded content
|
||||
new_chunks = Any[]
|
||||
new_chunks = []
|
||||
for c in chunks
|
||||
newc = CodeChunk(c.content, c.number, c.start_line, c.optionstring, c.options)
|
||||
newc.result_no = c.result_no
|
||||
|
|
|
|||
|
|
@ -1,23 +1,49 @@
|
|||
"""
|
||||
parse_markdown(document_body, is_pandoc = false)::Vector{WeaveChunk}
|
||||
parse_markdown(document_body, code_start, code_end)::Vector{WeaveChunk}
|
||||
|
||||
Parses Weave markdown and returns [`WeaveChunk`](@ref)s.
|
||||
"""
|
||||
function parse_markdown end
|
||||
|
||||
function parse_markdown(document_body, is_pandoc = false)::Vector{WeaveChunk}
|
||||
code_start, code_end = if is_pandoc
|
||||
r"^<<(?<options>.*?)>>=\s*$",
|
||||
r"^@\s*$"
|
||||
function parse_markdown(document_body; is_pandoc = false)
|
||||
if is_pandoc
|
||||
header = Dict()
|
||||
offset = 0
|
||||
code_start = r"^<<(?<options>.*?)>>=\s*$"
|
||||
code_end = r"^@\s*$"
|
||||
else
|
||||
r"^[`~]{3}(?:\{?)julia(?:;?)\s*(?<options>.*?)(\}|\s*)$",
|
||||
r"^[`~]{3}\s*$"
|
||||
header_text, document_body, offset = separate_header_text(document_body)
|
||||
header = parse_header(header_text)
|
||||
code_start = r"^[`~]{3}(?:\{?)julia(?:;?)\s*(?<options>.*?)(\}|\s*)$"
|
||||
code_end = r"^[`~]{3}\s*$"
|
||||
end
|
||||
return parse_markdown(document_body, code_start, code_end)
|
||||
return header, parse_markdown_body(document_body, code_start, code_end, offset)
|
||||
end
|
||||
|
||||
function parse_markdown(document_body, code_start, code_end)::Vector{WeaveChunk}
|
||||
# headers
|
||||
# -------
|
||||
|
||||
const HEADER_REGEX = r"^---$(?<header>((?!---).)+)^---$"ms
|
||||
|
||||
# TODO: non-Weave headers should keep live in a doc
|
||||
# separates header section from `text`
|
||||
function separate_header_text(text)
|
||||
m = match(HEADER_REGEX, text)
|
||||
isnothing(m) && return "", text, 0
|
||||
header_text = m[:header]
|
||||
return header_text, replace(text, HEADER_REGEX => ""; count = 1), count("\n", header_text)
|
||||
end
|
||||
|
||||
# HACK:
|
||||
# YAML.jl can't parse text including ``` characters, so first replace all the inline code
|
||||
# with these temporary code start/end string
|
||||
const HEADER_INLINE_START = "<weave_header_inline_start>"
|
||||
const HEADER_INLINE_END = "<weave_header_inline_end>"
|
||||
|
||||
function parse_header(header_text)
|
||||
isempty(header_text) && return Dict()
|
||||
pat = INLINE_REGEX => SubstitutionString("$(HEADER_INLINE_START)\\1$(HEADER_INLINE_END)")
|
||||
header_text = replace(header_text, pat)
|
||||
return YAML.load(header_text)
|
||||
end
|
||||
|
||||
# body
|
||||
# ----
|
||||
|
||||
function parse_markdown_body(document_body, code_start, code_end, offset)
|
||||
lines = split(document_body, '\n')
|
||||
|
||||
state = "doc"
|
||||
|
|
@ -25,7 +51,7 @@ function parse_markdown(document_body, code_start, code_end)::Vector{WeaveChunk}
|
|||
docno = 1
|
||||
codeno = 1
|
||||
content = ""
|
||||
start_line = 0
|
||||
start_line = offset
|
||||
|
||||
options = Dict()
|
||||
optionString = ""
|
||||
|
|
@ -57,7 +83,7 @@ function parse_markdown(document_body, code_start, code_end)::Vector{WeaveChunk}
|
|||
end
|
||||
|
||||
content = ""
|
||||
start_line = lineno
|
||||
start_line = lineno + offset
|
||||
|
||||
continue
|
||||
end
|
||||
|
|
@ -66,7 +92,7 @@ function parse_markdown(document_body, code_start, code_end)::Vector{WeaveChunk}
|
|||
chunk = CodeChunk(content, codeno, start_line, optionString, options)
|
||||
|
||||
codeno += 1
|
||||
start_line = lineno
|
||||
start_line = lineno + offset
|
||||
content = ""
|
||||
state = "doc"
|
||||
push!(chunks, chunk)
|
||||
|
|
|
|||
|
|
@ -1,9 +1,4 @@
|
|||
"""
|
||||
parse_notebook(document_body)::Vector{WeaveChunk}
|
||||
|
||||
Parses Jupyter notebook and returns [`WeaveChunk`](@ref)s.
|
||||
"""
|
||||
function parse_notebook(document_body)::Vector{WeaveChunk}
|
||||
function parse_notebook(document_body)
|
||||
nb = JSON.parse(document_body)
|
||||
chunks = WeaveChunk[]
|
||||
options = Dict{Symbol,Any}()
|
||||
|
|
@ -25,5 +20,5 @@ function parse_notebook(document_body)::Vector{WeaveChunk}
|
|||
end
|
||||
end
|
||||
|
||||
return chunks
|
||||
return Dict(), chunks
|
||||
end
|
||||
|
|
|
|||
|
|
@ -55,11 +55,8 @@ end
|
|||
function parse_doc(document, informat)
|
||||
document = replace(document, "\r\n" => "\n") # normalize line ending
|
||||
|
||||
header_text, document = separate_header_text(document)
|
||||
|
||||
return parse_header(header_text),
|
||||
informat == "markdown" ? parse_markdown(document) :
|
||||
informat == "noweb" ? parse_markdown(document, true) :
|
||||
return informat == "markdown" ? parse_markdown(document) :
|
||||
informat == "noweb" ? parse_markdown(document; is_pandoc = true) :
|
||||
informat == "script" ? parse_script(document) :
|
||||
informat == "notebook" ? parse_notebook(document) :
|
||||
error("unsupported input format given: $informat")
|
||||
|
|
@ -128,32 +125,6 @@ end
|
|||
|
||||
parse_inline(text) = Inline[InlineText(text, 1, length(text), 1)]
|
||||
|
||||
# headers
|
||||
# -------
|
||||
|
||||
const HEADER_REGEX = r"^---$(?<header>((?!---).)+)^---$"ms
|
||||
|
||||
# TODO: non-Weave headers should keep live in a doc
|
||||
# separates header section from `text`
|
||||
function separate_header_text(text)
|
||||
m = match(HEADER_REGEX, text)
|
||||
isnothing(m) && return "", text
|
||||
return m[:header], replace(text, HEADER_REGEX => ""; count = 1)
|
||||
end
|
||||
|
||||
# HACK:
|
||||
# YAML.jl can't parse text including ``` characters, so first replace all the inline code
|
||||
# with these temporary code start/end string
|
||||
const HEADER_INLINE_START = "<weave_header_inline_start>"
|
||||
const HEADER_INLINE_END = "<weave_header_inline_end>"
|
||||
|
||||
function parse_header(header_text)
|
||||
isempty(header_text) && return Dict()
|
||||
pat = INLINE_REGEX => SubstitutionString("$(HEADER_INLINE_START)\\1$(HEADER_INLINE_END)")
|
||||
header_text = replace(header_text, pat)
|
||||
return YAML.load(header_text)
|
||||
end
|
||||
|
||||
include("markdown.jl")
|
||||
include("script.jl")
|
||||
include("notebook.jl")
|
||||
|
|
|
|||
|
|
@ -1,9 +1,4 @@
|
|||
"""
|
||||
parse_script(document_body)::Vector{WeaveChunk}
|
||||
|
||||
Parse Julia script and returns [`WeaveChunk`](@ref)s.
|
||||
"""
|
||||
function parse_script(document_body)::Vector{WeaveChunk}
|
||||
function parse_script(document_body)
|
||||
lines = split(document_body, "\n")
|
||||
|
||||
doc_line = r"(^#'.*)|(^#%%.*)|(^# %%.*)"
|
||||
|
|
@ -20,8 +15,6 @@ function parse_script(document_body)::Vector{WeaveChunk}
|
|||
optionString = ""
|
||||
chunks = WeaveChunk[]
|
||||
state = "code"
|
||||
lineno = 1
|
||||
n_emptylines = 0
|
||||
|
||||
for lineno = 1:length(lines)
|
||||
line = lines[lineno]
|
||||
|
|
@ -81,12 +74,6 @@ function parse_script(document_body)::Vector{WeaveChunk}
|
|||
docno += 1
|
||||
end
|
||||
read *= line * "\n"
|
||||
|
||||
if strip(line) == ""
|
||||
n_emptylines += 1
|
||||
else
|
||||
n_emptylines = 0
|
||||
end
|
||||
end
|
||||
|
||||
# Handle the last chunk
|
||||
|
|
@ -98,5 +85,5 @@ function parse_script(document_body)::Vector{WeaveChunk}
|
|||
push!(chunks, chunk)
|
||||
end
|
||||
|
||||
return chunks
|
||||
return Dict(), chunks
|
||||
end
|
||||
|
|
|
|||
|
|
@ -17,14 +17,20 @@ ms = collect(eachmatch(Weave.INLINE_REGEXES, doc))
|
|||
@test ms[2][1] == "code"
|
||||
@test ms[3][1] == "show(\"is\")"
|
||||
|
||||
chunk = Weave.parse_markdown(doc)[1]
|
||||
let
|
||||
_, chunks = Weave.parse_markdown(doc)
|
||||
chunk = first(chunks)
|
||||
@test length(chunk.content) == 7
|
||||
@test chunk.content[2].content == ms[1][2]
|
||||
@test chunk.content[4].content == ms[2][1]
|
||||
@test chunk.content[6].content == ms[3][1]
|
||||
end
|
||||
|
||||
chunknw = Weave.parse_markdown(doc, false)[1]
|
||||
@test all([chunknw.content[i].content == chunk.content[i].content for i in 1:7])
|
||||
let
|
||||
_, chunks = Weave.parse_markdown(doc)
|
||||
chunk = first(chunks)
|
||||
@test all([chunk.content[i].content == chunk.content[i].content for i in 1:7])
|
||||
end
|
||||
|
||||
# Test with document
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue