Project File Options

Project File Options#

You can specify various options and information for your project in one of two ways:

  1. with a metadata block at the top of the project file;

  2. in the extra.ford table of your fpm.toml file (new in version 7.0).

fpm.toml File#

If you specify options in a fpm.toml file, you will still need a project file for the front page of your documentation. The main benefit of using fpm.toml is the improved semantics from the TOML format, making it easier to write lists and dicts (arrays and tables in TOML), as well as supporting comments.

Ford options should appear in the [extra.ford] table:

[extra.ford]
project = "My Project"

# TOML allows comments, which is handy
summary = """
Splitting your summary over multiple lines
is much easier in TOML.

It also allows blank lines in multi-line strings.
"""

preprocess = true
display = ["public", "protected"]
max_frontpage_items = 4
# An array of inline tables:
extra_filetypes = [
  { extension = "cpp", comment = "//" },
  { extension = "sh", comment = "#", lexer = "bash" },
  { extension = "py", comment = "#", lexer = "python" },
]

If an fpm.toml file with the [extra.ford] table exists in the same directory where you run ford, it will be automatically used and any metadata block in the project file will be ignored.

Note

If you use LaTeX in an option value, you probably want to use literal strings with single quotes to avoid errors about “Unknown escape character”

Markdown metadata#

Specifying options via the metadata block in the project file uses the same syntax as the Markdown Meta-Data extension, and consists of a single block at the top of the file, and ends with either a single blank line or triple-dashes ---.

Metadata consists of a series of keywords and values defined at the beginning of a markdown document like this:

---
project: My Project
summary: This is a summary of the project
    split over multiple lines (without any
    blank lines!)
preprocess: true
display: public
         protected
max_frontpage_items: 4
extra_filetypes: cpp //
    sh # bash
    py # python
---

This is the first paragraph of the document.

The keywords are case-insensitive and may consist of letters, numbers, underscores and dashes and must end with a colon. The values consist of anything following the colon on the line and may even be blank.

If a line is indented by 4 or more spaces, that line is assumed to be an additional line of the value for the previous keyword. A keyword may have as many lines as desired – note that these must be spaces and not tabs. Also note that you cannot have any blank lines in a list item or they will be interpreted as closing the metadata block.

The first blank line ends all metadata for the document. Therefore, the first line of a document must not be blank. All metadata is stripped from the document prior to any further processing by Markdown.

Project file options will be overriden by command line options .See ./example/example-project-file.md for a sample project file.

Except where noted, all paths in options are interpreted relative to the path of the project file.

Note

Markdown comments must not appear within the meta data section! Typical markdown commenting strategies may be used within the markdown body of the project file, BUT NOT WITHIN THE META-DATA SECTION! After declaring metadata HTML block comments of the form

<!-- This is a multi line
comment!

wow
-->

or markdown phony link comments may be used:

[comment 1 goes here, this will declare a phony link target. Just make sure not to reference the null anchor]:#

Unless stated, most options are plain strings:

doc_license = "by-sa"

doc_license: by-sa

If an option is a list, it can take multiple values. Note that for markdown-metadata, each value must be on its own line and indented by at least four spaces:

macro = [
  "HAS_DECREMENT",
  "DIMENSION=3",
]

macro: HAS_DECREMENT
    DIMENSION=3

Project Information#

Information about your project.

doc_license#

The license under which the documentation is released. See license for possible values.

favicon#

The path to a custom favicon which will be used by the HTML documentation. If left blank, it will default to an icon for FORD.

gitter_sidecar#

The name of the project’s chatroom on Gitter, which can then be displayed using the Gitter sidecar.

license#

The licenses under which the software is released. Options are:

privacy_policy_url#

URL of the privacy policy of the project.

project#

The name of this project. (default: Fortran Project)

project_bitbucket#

The URL of the BitBucket repository for this project.

project_download#

A URL from which to download the source or binaries for this project.

project_github#

The URL of the Github repository for this project.

project_gitlab#

The URL of the Gitlab repository for this project.

project_sourceforge#

The Sourceforge repository for this project.

project_url#

The URL at which the documentation will be available. If left blank then relative URLs will be used for links. This can be used within any documentation with the macro |url|. (default: blank, i.e. relative links)

project_website#

The homepage for this project.

summary#

A summary of the description of your project. If present it will be printed in a “Jumbotron” element at the top of the documentation index page. This will be processed by Markdown before being used.

terms_of_service_url#

URL of the terms of service of the project

Author Information#

Information about the author.

author#

The name of the person(s) or organization who wrote this project.

author_description#

A brief description of the author. You could provide biographical details or links to other work, for example. This will be processed by Markdown before being used.

author_pic#

A picture of or avatar for the author.

bitbucket#

The author’s BitBucket page.

email#

The author’s email address.

facebook#

The author’s Facebook profile.

github#

The author’s Github page.

gitlab#

The author’s Gitlab page.

google_plus#

The author’s Google+

linkedin#

The author’s LinkedIn profile.

twitter#

The author’s Twitter.

website#

The author’s website.

Directories#

Settings specifying where to look (and not to look) for documentation.

copy_subdir#

A list of subdirectories to copy verbatim into the generated documentation. See Copying directories: copy_subdir for a more detailed explanation of this option. (optional)

exclude_dir#

List of directories whose contents should not be included in documentation. Provide the relative path to directory from the top level project file. Can be a glob pattern, for example **/test*, which will match any directory that starts with test anywhere in the source directory tree.

exclude_dir = [
  "**/test*",
  "src/internal",
]

exclude_dir: **/test*
    src/internal

html_template_dir#

A list of directories to search for HTML templates.

Caution

This is an experimental feature!

You will most likely want to copy-paste the existing templates and modify them. They are not well documented, so use at your own risk!

include#

Directories in which the C preprocessor searches for any #includeed files, such as headers. These directories will also be searched for files loaded using Fortran’s intrinsic include statement.

media_dir#

A directory containing any images or other content which you will use or link to in your documentation. This will be placed at the root of your documentation file-tree, with the name “media”. The URL of this directory can be accessed within your documentation using the macro |media|.

md_base_dir#

The directory relative to which any “included” Markdown files’ paths are specified. (default: directory containing the project file.)

page_dir#

A directory containing markdown files to be processed into individuals pages within the documentation. See Static Pages for details.

src_dir#

List of directories where the source-files are to be found for this project. These must not be a subdirectory of the output_dir (see below). (default: ["./src"])

Source File Settings#

Settings related to individual source files.

encoding#

The text encoding to use when opening source files (default: utf-8)

exclude#

List of source files which should not be included in documentation. This should either be a relative path that includes one of the source directories, or a glob pattern. For example, src/not_this.f90 to exclude a specific file, or **/test_*.f90 to exclude any .f90 files that start with test_ anywhere in any of the source directories.

Deprecated since version 7.0.0: In earlier versions, not_this.f90 would exclude any file called not_this.f90 anywhere in the project. This will now emit a warning, and should be changed to either a relative path (src/not_this.f90) or a glob pattern (**/not_this.f90)

exclude = [
  "**/test_*.F90",
  "src/generated_file.f90",
]

exclude: **/test_*.F90
    src/generated_file.f90

extensions#

List of file extensions (without the dot) which will be read by FORD for documentation. These extensions are only for free-form code; see fixed_extensions for fixed-form extensions. (default: f90, f95, f03, f08, f15, F90, F95, F03, F08, F15)

extensions = ["f90", "f", "F90", "F"]

extensions: f90
    f
    F90
    F

extra_filetypes#

List of non-Fortran filetypes from which documentation should be extracted (see Non-Fortran Source Files).

Experimental: You may optionally specify the Pygments lexer to use when applying syntax-highlighting to the file, as an additional argument after the comment character. This should take the form of the module being imported relative to pygments.lexer, e.g. fortran.FortranLexer or c_cpp.CLexer. This feature should not be considered stable and the behaviour may change in future releases. If you don’t supply this, pygments will guess which lexer to use based the file extension and some lexical analysis.

For TOML config files, this should be a list of tables with required keys extension and comment, and an optional lexer.

For Markdown metadata config, each entry must be on its own line and should consist of the filetype extension, a space, and then the character(s) designating a comment. Only single-line comments are supported.

extra_filetypes = [
  { extension = "cpp", comment = "//" },
  { extension = "sh", comment = "#", lexer = "bash" },
  { extension = "py", comment = "#", lexer = "python" },
]

extra_filetypes: cpp //
    sh # bash
    py # python

fixed_extensions#

List of file extensions which will be read by FORD for documentation, with the files containing fixed-form code. (default: f, for, F, FOR)

fixed_length_limit#

If false, fixed-form code lines are read in their entire length. Otherwise anything after the 72nd column is ignored. (default: true)

Preprocessing#

If desired, your source files can be passed through an arbitrary preprocessor before being analysed by FORD.

fpp_extensions#

File extensions which should be preprocessed prior to further analysis. If the extension is not specified in extensions or fixed_extensions then the file will be assumed to be free-form. (default: F90, F95, F03, F08, F15, F, FOR)

macro#

List of macros to be provided to the C preprocessor when applying it to source files. Can take the form mac-name or mac-name=mac-value.

macro = [
  "HAS_DECREMENT",
  "DIMENSION=3",
]

macro: HAS_DECREMENT
    DIMENSION=3

preprocess#

If set to ‘true’, then any files with extensions in fpp_extensions will be passed through the specified preprocessor, CPP by default. (default: true)

preprocessor#

The preproccessor command to use on files with extensions in fpp_extensions. Can include flags as needed. Preprocessor macros and include paths specified in the project file will automatically be appended using the CPP interface, which is fairly standard. (default: pcpp -D__GFORTRAN__)

Documentation Markers#

docmark#

The symbol(s) following an exclaimation mark which designates that a comment contains documentation. For example, if the docmark was !, comments would then be designated by !!. It should not be the same as any other docmark. (default: !)

docmark_alt#

The symbol(s) following an exclaimation mark which designate that the following set of comments, until the end of the block, are all documentation. This mark needs only to be used at the beginning of the block, after which all regular comments will be treated as documentation. For example, if the docmark was *, comments would then be designated by !*. An example of such a block of documentation is provided.

!* This is an example.
!  Here is another line of comments.
!
!  History
! ----------
!  * 1/1/2000 Created

subroutine blah()

end subroutine blah

It should not be the same symbol as any other docmark. (default: *)

predocmark#

The symbol(s) following an exclaimation mark which designates that a comment contains documentation preceding the code which it is documenting. For example, if the docmark was >, comments would then be designated by !>. It should not be the same as any other docmark. (default: >)

predocmark_alt#

The symbol(s) following an exclaimation mark which designate the start of a block of documentation preceding the code which it is documenting and that all further comments within this block will be treated as documentation. For example, if the predocmark_alt was #, comments would then be designated by !#. It should not be the same as any other docmark. (default: |)

Documentation Settings#

Settings specifying how to process documentation and what information to display in the output.

alias#

List of aliases in the form key = replacement. In the documentation |key| can then be used as shorthand for replacement. For example:

# As an in-line table (note that this has to all be on one line!)
alias = {ford = "FORD (the Fortran documentation generator", euler = '\exp(i \pi) + 1 = 0'}

# Or as a separate table:
[extra.ford.alias]
ford = "FORD (the Fortran documentation generator"
euler = '\exp(i \pi) + 1 = 0'

alias: ford = FORD (the Fortran documentation generator)
       euler = \exp(i \pi) + 1 = 0

and the markdown:

This code uses |ford|.
Did you know Euler's identity is $$|euler|$$?

becomes:

This software uses FORD (the Fortran documentation generator).
Did you know Euler's identity is $$\exp(i \pi) + 1 = 0$$?

Three aliases are pre-defined:

  • |url| for the project URL,

  • |media| for the media directory, and

  • |page| for the page_dir.

Note:#

Aliases can currently only be defined in the project file, and not in individual docstrings

creation_date#

A Python datetime format to be used if the date and time at which the documentation was generated is printed in the documentation footer. (default: %Y-%m-%dT%H:%M:%S.%f%z)

css#

The path to a custom style-sheet which can be used to modify the appearance of the output.

display#

How much documentation should be printed. Options are ‘public’, ‘private’, ‘protected’, or any combination of those three. If ‘none’ is present, then nothing will be displayed other than the programs, modules, and procedures contained within source files (that is, procedures within modules will not be displayed). These choice can be overridden for a specific item using the documentation meta data, and those settings will be inherited by any items they contain. (default: ‘public’ and ‘protected’)

external#

Paths or URLs of external projects to link to. If an entity is not found in the sources, FORD will try to look it up in those external projects. If those have documentation generated by FORD with the externalize option, a link will be placed into the documentation wherever this entity is referred to. FORD will look in the provided paths for a modules.json file.

The difference between external between extra_mods is that FORD can link directly to entities (functions, types, and so on) with external, while only modules will be linked to using extra_mods.

extra_mods#

A list of modules (and their external documentation) which are not included in the project. For the markdown metadata, an entry takes the form module_name:url where module_name is its name as it would appear in a use statement, and url is the location of its documentation. For TOML options, use a table with module_name = "url" as the key-value pairs. Any entity which uses this module will provide a link to the external documentation in the same way that it would provide a link to the documentation of a module in the project.

# As an inline table:
extra_mods = { example_mod = "https://example.com" }

# Or you might find it easier as a separate table:
[extra.ford.extra_mods]
iso_fortran_env = "https://gcc.gnu.org/onlinedocs/gfortran/ISO_005fFORTRAN_005fENV.html"
example_mod = "https://example.com"

extra_mods: iso_fortran_env: "https://gcc.gnu.org/onlinedocs/gfortran/ISO_005fFORTRAN_005fENV.html"
      example_mod: https://example.com

extra_vartypes#

A list of extra types of variables which FORD should look for. This can be useful when using, for example, the PETSc library.

hide_undoc#

If true, then don’t display any undocumented entities (default: false)

incl_src#

This flag toggles visibility of the source files in FORD documentation output. If set to true, the individual files will be listed and all contents will be shown on a file page. If false, procedures will still show the names of the files they are defined in, but there will be no way to access the contents of the file itself. For showing the code definitions for individual procedures, modules, and derived types, see source. (default: true)

lower#

If true then convert all non-string and non-comment source code to lower case prior to analyzing. (default: false)

mathjax_config#

The path to a JavaScript file containing settings for MathJax. This might be used to, e.g., define TeX macros.

max_frontpage_items#

The maximum number of items to list under each category of entity on the front page. (default: 10)

md_extensions#

List of Markdown extensions which you wish to be used when parsing your documentation. For example, markdown.extensions.toc. Note that Markdown-Extra, CodeHilite, and Meta-Data are loaded by default.

proc_internals#

If false then the local variables, derived types, etc. within public procedures will not be included in documentation. This is equivalent to setting display: none in the documentation meta data of each procedure. It can be overriden locally in the documentation meta data. (default: false)

revision#

The name of the particular revision of your code/documentation, to be printed in the footer below the license and copyright year.

sort#

The order in which to display entities (variables, procedures, etc.) in the documentation. Options are (default: src)

  • src: Order which they occur in source code

  • alpha: Alphabetical order

  • permission: Display public first, then protected, then private. Within these categories, items are displayed in the same order as they occur in the source code.

  • permission-alpha: Display public first, then protected, then private. Within these categories, items are displayed in alphabetical order.

  • type: Sort variables (and functions) by type. For each time, items are displayed in the same order as they occur in the source code

  • type-alpha: Sort variables (and functions) by type. Within these categories, items are displayed in alphabetical order.

source#

If set to ‘true’, then the syntax-highlighted source code will be displayed at the bottom of the documentation page for each procedure, program, and derived type. This behaviour can be overridden for a given item using the documentation meta data. FORD may not be able to extract the source code in all cases; see source for details. To hide source files themselves, see incl_src. Note that this substantially increases run-time. (default: false)

version#

The version name/number of your project.

year#

The year to display in the copyright notice. (default: the current year)

Graph Settings#

FORD can generate call-trees, dependency diagrams, and inheritance diagrams will be produced for the project. These require Graphviz to be installed. Note that this can increase run-time substantially. The following graphs are produced: - For each module: - a graph showing the modules which it uses and, if a submodule, the (sub)modules it is descended from - a graph showing which modules use and which submodules descend from this one - For each type: - a graph showing all type which it descends from or contains as a component - a graph showing all types which descend from or contain as a component this type - For each procedure: - a graph showing all procedures called by this procedure and (for interfaces) any procedures which it provides an interface to - a graph showing all procedures which call this one or provide and interface to it - For each program: - a graph showing the modules which are used by the program - a graph showing the procedures called by the program - A graph showing all module use dependencies on the module list page - A graph showing the inheritance structure of all derived types (and their use as components of other types) on the type list page - A graph showing the call-tree for all programs and procedures on the procedure list page

Note that, at present, call-trees only work for procedural programming and will not identify any calls to type-bound procedures. Call-trees are not supposed to show intrinsic procedures. However, intrinsic procedures and even keywords may appear in a grey node on the graph. This means that it this procedure was not known (or overlooked) by the developers. Please report this is a bug. (default: false)

coloured_edges#

If true then edges connecting nodes in the graphs will be assigned various colours. This can make large graphs easier to read. Americans, please note that the proper spelling has been used here. (default: false)

graph#

If set to ‘true’ then graphs are produced of call trees, dependency structures, and inheritance diagrams. This behaviour can be overridden for a given item in the code using the documentation meta data. (default: false)

graph_maxdepth#

The maximum number of recursions to make when analysing graph structures. For large projects, producing graphs can be prohibitively time-consuming and the graphs confusing and unreadable if full recursion is used, so you may wish to set the maximum to be only a few levels. (default: 10000)

graph_maxnodes#

The maximum number of nodes which may be displayed in a graph. For large projects, graphs become unreadable if they contain too many nodes. A graph’s depth will be reduced to keep the number of nodes below this maximum or, if the even a depth of one would result in more nodes than the maximum, it will be restructured to give a clearer visualisation. (default: 100000000)

show_proc_parent#

If true then the parent module of a procedure will be displayed in the graphs as follows: parent::procedure. (default: false)

Output#

Where documentation should be written to.

externalize#

Create a modules.json file under output_dir containing information about entities and the URL of their documentation. This allows this project to be used as an external link in another project.

graph_dir#

A directory where, if it is specified and graphs is set to true, SVG and graphviz copies of all graphs for your project will be placed. Note that name mangling is applied to the filenames.

output_dir#

The directory where the project output will be placed. Any content already present there will be deleted. (default: ./doc)

Run-Time Behaviour#

Miscellaneous options determining how FORD is run and its output.

dbg#

Allows FORD to crash and display a Python backtrace if an error is encountered when parsing a file.

force#

Try to continue as much as possible, even if there are fatal errors when reading files.

parallel#

The number of CPUs to in multithreading. 0 indicates that the code should be run in serial. (default: number of cores on the computer)

quiet#

If ‘true’, FORD will suppress all output documenting its progress. (default: false)

warn#

If ‘true’, FORD will print warning messages for any undocumented items which it encounters and any time it can not find the source code for some item where it is requested as part of the documentation. (default: false)