Ordered snippets with file system hierarchy

This commit is contained in:
Felix Krause 2016-10-08 20:57:53 +02:00
parent 31ac201e41
commit 9f7e87e935
37 changed files with 116 additions and 133 deletions

View File

@ -1,43 +0,0 @@
{
"Dumping Nim objects as YAML": [
"dump-yaml", "out"
],
"Loading Nim objects from YAML": [
"load-yaml", "in"
],
"Customizing output style" : [
"outputstyle", "out"
],
"Dumping reference types and cyclic structures": [
"dump-reftypes", "out"
],
"Loading reference types and cyclic structures": [
"load-reftypes", "in"
],
"Defining a custom tag uri for a type": [
"customtag", "out"
],
"Dumping Nim objects as JSON": [
"dump-json", "out"
],
"Loading Nim objects from JSON": [
"load-json", "in"
],
"Processing a Sequence of Heterogeneous Items": {
"… With variant objects": [
"implicit-variant", "in"
],
"… With the Sequential API": [
"sequential-api", "in"
]
}
}

View File

@ -16,7 +16,4 @@ install it with `Nimble <https://github.com/nim-lang/nimble>`_:
.. code-block:: bash
nimble install yaml
Quickstart
==========
%examples/quickstart.json%/
%quickstart%0

View File

@ -1,9 +1,10 @@
## This is a tool for preprocessing rst files. Lines starting with ``%`` will
## get substituted by nicely layouted included nim and yaml code.
## get substituted by nicely layouted nim and yaml code included from file in
## the snippets tree.
##
## The syntax of substituted lines is ``'%' jsonfile '%' jsonpath``. *jsonfile*
## shall be the path to a JSON file. *jsonpath* shall be a path to some node in
## that JSON file.
## The syntax of substituted lines is ``'%' path '%' level``. *path* shall be
## a path relative to the *snippets* directory. *level* shall be the level depth
## of the first title that should be produced.
##
## Usage:
##
@ -11,8 +12,12 @@
##
## *path* is the output path. If omitted, it will be equal to infile with its
## suffix substituted by ``.rst``. *infile* is the source rst file.
##
## The reason for this complex approach is to have all snippets used in the docs
## available as source files for automatic testing. This way, we can make sure
## that the code in the docs actually works.
import parseopt2, json, streams, tables, strutils, os
import parseopt2, streams, tables, strutils, os
var
infile = ""
@ -57,71 +62,73 @@ var tmpOut = newFileStream(path, fmWrite)
proc append(s: string) =
tmpOut.writeLine(s)
proc gotoPath(root: JsonNode, path: string): JsonNode =
doAssert path[0] == '/'
if path.len == 1: return root
doAssert root.kind == JObject
for i in 1..<path.len:
if path[i] == '/':
return gotoPath(root.getFields()[path[1..<i]], path[i+1 .. ^1])
return root.getFields()[path[1..^1]]
const headingChars = ['=', '-', '`', ':', '\'']
const headingChars = ['-', '`', ':', '\'']
proc outputExamples(node: JsonNode, prefix: string, level: int = 0) =
case node.kind
of JObject:
for key, value in node.getFields():
append(key)
proc outputExamples(curPath: string, level: int = 0) =
let titlePath = curPath / "title"
if fileExists(titlePath):
let titleFile = open(titlePath, fmRead)
defer: titleFile.close()
var title = ""
if titleFile.readLine(title):
let headingChar = if level >= headingChars.len: headingChars[^1] else:
headingChars[level]
append(repeat(headingChar, key.len) & '\l')
outputExamples(value, prefix, level + 1)
of JArray:
let elems = node.getElems()
case elems.len
of 2:
append(".. raw:: html")
append(" <table class=\"quickstart-example\"><thead><tr><th>code.nim</th>")
append(" <th>" & elems[1].getStr() &
".yaml</th></tr></thead><tbody><tr><td>\n")
append(".. code:: nim")
append(" :file: " & prefix & elems[0].getStr() & ".nim\n")
append(".. raw:: html")
append(" </td>\n <td>\n")
append(".. code:: yaml")
append(" :file: " & prefix & elems[0].getStr() & '.' &
elems[1].getStr() & ".yaml\n")
append(".. raw:: html")
append(" </td></tr></tbody></table>\n")
else:
echo "Unexpected number of elements in array: ", elems.len
quit 1
append(title)
append(repeat(headingChar, title.len) & '\l')
# process content files under this directory
var codeFiles = newSeq[string]()
for kind, filePath in walkDir(curPath, true):
if kind == pcFile:
if filePath != "title": codeFiles.add(filePath)
case codeFiles.len
of 0: discard
of 1:
let (nullPath, name, extension) = codeFiles[0].splitFile()
append(".. code:: " & extension[1..^1])
append(" :file: " & (curPath / codeFiles[0]) & '\l')
of 2:
append(".. raw:: html")
append(" <table class=\"quickstart-example\"><thead><tr>")
for codeFile in codeFiles:
append(" <th>" & codeFile & "</th>")
append(" </th></tr></thead><tbody><tr><td>\n")
var first = true
for codeFile in codeFiles:
if first: first = false
else: append(".. raw:: html\n </td>\n <td>\n")
let (nullPath, name, extension) = codeFile.splitFile()
append(".. code:: " & extension[1..^1])
append(" :file: " & (curPath / codeFile) & '\l')
append(".. raw:: html")
append(" </td></tr></tbody></table>\n")
else:
echo "Unexpected node kind: ", node.kind
quit 1
echo "Unexpected number of files in ", curPath, ": ", codeFiles.len
# process child directories
for kind, dirPath in walkDir(curPath):
if kind == pcDir:
outputExamples(dirPath, level + 1)
var lineNum = 0
for line in infile.lines():
if line.len > 0 and line[0] == '%':
var
jsonFile: string = nil
jsonPath: string = nil
srcPath: string = nil
level = 0
for i in 1..<line.len:
if line[i] == '%':
jsonFile = line[1 .. i - 1]
jsonPath = line[i + 1 .. ^1]
srcPath = line[1 .. i - 1]
level = parseInt(line[i + 1 .. ^1])
break
if isNil(jsonFile):
if isNil(srcPath):
echo "Second % missing in line " & $lineNum & "! content:\n"
echo line
quit 1
let root = parseFile(jsonFile)
var prefix = ""
for i in countdown(jsonFile.len - 1, 0):
if jsonFile[i] == '/':
prefix = jsonFile[0..i]
break
outputExamples(root.gotoPath(jsonPath), prefix)
outputExamples("snippets" / srcPath, level)
else:
append(line)

View File

@ -0,0 +1 @@
Dumping Nim objects as YAML

View File

@ -0,0 +1 @@
Loading Nim objects from YAML

View File

@ -0,0 +1 @@
Customizing output style

View File

@ -0,0 +1 @@
Dumping reference types and cyclic structures

View File

@ -0,0 +1 @@
Loading reference types and cyclic structures

View File

@ -0,0 +1 @@
Defining a custom tag uri for a type

View File

@ -0,0 +1 @@
Dumping Nim objects as JSON

View File

@ -0,0 +1 @@
Loading Nim objects from JSON

View File

@ -0,0 +1 @@
… With variant objects

View File

@ -0,0 +1 @@
… With the Sequential API

View File

@ -0,0 +1 @@
Processing a Sequence of Heterogeneous Items

View File

@ -0,0 +1 @@
Quickstart

View File

@ -738,15 +738,20 @@ macro constructImplicitVariantObject(s, c, r, possibleTagIds: untyped,
))
ifStmt.add(newNimNode(nnkElse).add(newNimNode(nnkTryStmt).add(
newStmtList(raiseStmt), newNimNode(nnkExceptBranch).add(
newIdentNode("KeyError"), newStmtList(newCall("internalError",
newStrLitNode("Unexcpected KeyError")))
newNimNode(nnkDiscardStmt).add(newEmptyNode())
))))
result = newStmtList(newCall("reset", r), ifStmt)
let implicitVariantObjectMarker {.compileTime.} =
newIdentNode(":implicitVariantObject")
macro isImplicitVariantObject(o: typed): untyped =
result = newCall("compiles", newCall(implicitVariantObjectMarker, o))
proc constructChild*[T](s: var YamlStream, c: ConstructionContext,
result: var T) =
let item = s.peek()
when compiles(implicitVariantObject(result)):
when isImplicitVariantObject(result):
var possibleTagIds = newSeq[TagId]()
case item.kind
of yamlScalar:
@ -915,7 +920,7 @@ proc representChild*[O](value: ref O, ts: TagStyle, c: SerializationContext) =
proc representChild*[O](value: O, ts: TagStyle,
c: SerializationContext) =
when compiles(implicitVariantObject(value)):
when isImplicitVariantObject(value):
# todo: this would probably be nicer if constructed with a macro
var count = 0
for name, field in fieldPairs(value):
@ -1032,3 +1037,34 @@ proc dump*[K](value: K, tagStyle: TagStyle = tsRootOnly,
try: result = present(events, serializationTagLibrary, options)
except YamlStreamError:
internalError("Unexpected exception: " & getCurrentException().repr)
proc canBeImplicit(t: typedesc): bool {.compileTime.} =
let tDesc = getType(t)
if tDesc.kind != nnkObjectTy: return false
if tDesc[2].len != 1: return false
if tDesc[2][0].kind != nnkRecCase: return false
var foundEmptyBranch = false
for i in 1.. tDesc[2][0].len - 1:
case tDesc[2][0][i][1].len # branch contents
of 0:
if foundEmptyBranch: return false
else: foundEmptyBranch = true
of 1: discard
else: return false
return true
macro setImplicitVariantObjectMarker(t: typedesc): untyped =
result = quote do:
proc `implicitVariantObjectMarker`*(unused: `t`) = discard
template markAsImplicit*(t: typedesc): typed =
## Mark a variant object type as implicit. This requires the type to consist
## of nothing but a case expression and each branch of the case expression
## containing exactly one field - with the exception that one branch may
## contain zero fields.
when canBeImplicit(t):
# this will be checked by means of compiles(implicitVariantObject(...))
setImplicitVariantObjectMarker(t)
else:
{. fatal: "This type cannot be marked as implicit" .}

View File

@ -253,32 +253,6 @@ template setTagUri*(t: typedesc, uri: string, idName: untyped): typed =
proc yamlTag*(T: typedesc[t]): TagId {.inline, raises: [].} = idName
## autogenerated
proc canBeImplicit(t: typedesc): bool {.compileTime.} =
let tDesc = getType(t)
if tDesc.kind != nnkObjectTy: return false
if tDesc[2].len != 1: return false
if tDesc[2][0].kind != nnkRecCase: return false
var foundEmptyBranch = false
for i in 1.. tDesc[2][0].len - 1:
case tDesc[2][0][i][1].len # branch contents
of 0:
if foundEmptyBranch: return false
else: foundEmptyBranch = true
of 1: discard
else: return false
return true
template markAsImplicit*(t: typedesc): typed =
## Mark a variant object type as implicit. This requires the type to consist
## of nothing but a case expression and each branch of the case expression
## containing exactly one field - with the exception that one branch may
## contain zero fields.
when canBeImplicit(t):
# this will be checked by means of compiles(implicitVariantObject(...))
proc implicitVariantObject*(unused: t) = discard
else:
{. fatal: "This type cannot be marked as implicit" .}
static:
# standard YAML tags used by serialization
registeredUris.add("!")