| Title: | Categorized views of R package repositories |
|---|---|
| Description: | Infrastructure to support 'views' used to classify Bioconductor packages. 'biocViews' are directed acyclic graphs of terms from a controlled vocabulary. There are three major classifications, corresponding to 'software', 'annotation', and 'experiment data' packages. |
| Authors: | VJ Carey <[email protected]>, BJ Harshfield <[email protected]>, S Falcon <[email protected]> , Sonali Arora, Lori Shepherd <[email protected]> |
| Maintainer: | Bioconductor Package Maintainer <[email protected]> |
| License: | Artistic-2.0 |
| Version: | 1.75.0 |
| Built: | 2024-11-29 03:53:26 UTC |
| Source: | https://github.com/bioc/biocViews |
Structures for vocabularies and narratives of views. This can be used to create HTML views of the package structure in a Bioconductor repository.
| Package: | biocViews |
| Version: | 1.11.4 |
| Depends: | R (>= 2.4.0), methods, utils |
| Imports: | tools, Biobase, graph (>= 1.9.26), RBGL (>= 1.13.5), XML |
| Suggests: | Biobase |
| License: | Artistic-2.0 |
| URL: | http://www.bioconductor.org/packages/release/BiocViews.html |
| biocViews: | Infrastructure |
Index:
BiocView-class Class "BiocView"
Htmlized-class Class "Htmlized"
PackageDetail-class Class "PackageDetail"
RepositoryDetail-class
Class "RepositoryDetail"
biocViewsVocab Bioconductor Task Views Vocabulary Data
extractVignettes Extract pdf vignettes from local package
repository
genReposControlFiles Generate CRAN-style repository control files
getBiocSubViews Build a list of BiocView objects from a package
repository
getBiocViews Build a list of BiocView objects from a package
repository
getPacksAndViews Parse VIEWS file for views and packages
getSubTerms Retrieve a term and its children from a vocab
DAG
htmlDoc Create a complete HTML document representation
of an object
htmlFilename Return a filename for an object's HTML
representation
htmlValue HTML Representation of an Object
writeBiocViews Write a list of BiocView objects to HTML
writeHtmlDoc Write an XML DOM containing HTML to a file
writePackageDetailHtml
Write HTML files for packages in a CRAN-style
repository
writeRepositoryHtml Write package descriptions and a repository
index as HTML
writeTopLevelView Write the view for the root of a vocabulary to
disk
write_REPOSITORY Write a REPOSITORY control file for a
CRAN-style package repository
write_SYMBOLS Write a SYMBOLS file
write_VIEWS Write a VIEWS control file for a CRAN-style
package repository
The terms of the vocabulary are stored in a DAG, which can be loaded as
the serialized data object biocViewsVocab. For listing of
available terms use function getSubTerms.
Further information is available in the following two vignettes:
HOWTO-BCV |
Basic package usage |
createReposHtml |
Further information for repository admins |
VJ Carey <[email protected]>, BJ Harshfield <[email protected]>, S Falcon <[email protected]>
Maintainer: Biocore Team c/o BioC user list <[email protected]>
data(biocViewsVocab) getSubTerms(biocViewsVocab, "Technology")data(biocViewsVocab) getSubTerms(biocViewsVocab, "Technology")
Representation of of Bioconductor "view".
Objects can be created by calls of the form new("BiocView", ...).
name:Object of class "character" giving the
name of the view.
subViews:Object of class "character" giving
the names of the subviews of this view.
parentViews:Object of class "character" giving
the names of the views that are this view's parents.
Title:Object of class "character" giving
longer description of view?
reposRoot:Object of class "character" URL for repository
homeUrl:Object of class "character" ?
htmlDir:Object of class "character" ?
packageList:Object of class "list" consisting
of PackageDetail-class objects
Class "RepositoryDetail", directly.
Class "Htmlized", directly.
signature(from = "BiocView", to = "rdPackageTable"): ...
signature(object = "BiocView"): ...
signature(object = "BiocView"): ...
signature(object = "BiocView"): ...
signature(object = "BiocView"): ...
Seth Falcon
A graphNEL-class instance representing the Bioconductor
Task Views as a directed graph.
data(biocViewsVocab)data(biocViewsVocab)
The format is: graphNEL instance
The source for the vocabulary data is in the dot directory of the
package in file biocViewsVocab.dot. This is transformed to GXL using
the dot2gxl command line utility from the graphviz package. Then the
fromGXL function from the graph package is used to
convert to graphNEL-class.
data(biocViewsVocab) biocViewsVocab ## If you have Rgraphviz available, you can ## plot the vocabulary with plot(biocViewsVocab)data(biocViewsVocab) biocViewsVocab ## If you have Rgraphviz available, you can ## plot the vocabulary with plot(biocViewsVocab)
produce character stream for use with Protege entities/class hierarchy ingestion based on biocViews subgraphs
dump_concept(x, edob = edges(biocViewsVocab))dump_concept(x, edob = edges(biocViewsVocab))
x |
character(1) name of a node in the biocViewsVocab graph |
edob |
a list as produced by graph::edges |
a character atom with x abutting left and edge targets tab-indented
cat(substring(dump_concept("ResearchField"), 1, 152), "\n")cat(substring(dump_concept("ResearchField"), 1, 152), "\n")
This function extracts Rd man pages and builds pdf reference manuals
from the man subdirectory of R source packages archives
(.tar.gz) found in a local package repository.
All Rd files found in man will be extracted and used during
the pdf construction process. Only source package archives will be
processed. The constructed pdf files will be extracted under
destDir and will be found in PKGNAME/man/*.pdf.
Prior to extraction, all Rd and pdf files in destDir/PKGNAME/man
will be removed.
extractManuals(reposRoot, srcContrib, destDir)extractManuals(reposRoot, srcContrib, destDir)
reposRoot |
character vector giving the path to the root of the local CRAN-style package repository |
srcContrib |
character vector giving the relative path from the
|
destDir |
character vector specifying the directory in which the
extracted files will be written. If missing, files will be written
to |
Patrick Aboyoun
Extracts NEWS files from source tarballs of packages.
extractNEWS(reposRoot, srcContrib, destDir)extractNEWS(reposRoot, srcContrib, destDir)
reposRoot |
Top level path for CRAN-style repos |
srcContrib |
Location of source packages |
destDir |
where to extract |
Extracts files from source tarballs of packages.
extractTopLevelFiles(reposRoot, srcContrib, destDir, fileName)extractTopLevelFiles(reposRoot, srcContrib, destDir, fileName)
reposRoot |
Top level path for CRAN-style repos |
srcContrib |
Location of source packages |
destDir |
where to extract |
fileName |
name of file to extract |
These functions extract pdf or HTML
files from the inst/doc subdirectory
of R source packages archives (.tar.gz) found in a local
package repository.
All pdf files found in inst/doc will be extracted.
With extractHTMLDocuments, all HTML files except
index.html will be extracted.
Only source
package archives will be processed. The extracted pdf
or HTML files will be
extracted under destDir and will be found in
PKGNAME/inst/doc/.
Prior to extraction, all pdf files in destDir/PKGNAME/inst/doc
will be removed.
extractVignettes(reposRoot, srcContrib, destDir) extractHTMLDocuments(reposRoot, srcContrib, destDir)extractVignettes(reposRoot, srcContrib, destDir) extractHTMLDocuments(reposRoot, srcContrib, destDir)
reposRoot |
character vector giving the path to the root of the local CRAN-style package repository |
srcContrib |
character vector giving the relative path from the
|
destDir |
character vector specifying the directory in which the
extracted files will be written. If missing, files will be written
to |
Seth Falcon
This function generates control files for CRAN-style repositories.
For each path specified in contribPaths a PACKAGES file
is written. In addition, two top-level control files are created:
REPOSITORY contains information about the specified contrib
paths.
VIEWS contains metadata for all packages in the repository
including the paths to any extracted vignettes, if found. This file
is useful for generating HTML views of the repository.
genReposControlFiles(reposRoot, contribPaths, manifestFile = NA, meatPath = NA)genReposControlFiles(reposRoot, contribPaths, manifestFile = NA, meatPath = NA)
reposRoot |
character vector containing the path to the CRAN-style repository root directory. |
contribPaths |
A named character vector. Valid names are
|
manifestFile |
character(1). File path location to Bioconductor formatted manifest file that lists all current packages. This file will be used in the write_VIEWS function to cross check successfully built packages with all expected packages. Packages that have not built will be given dummy entry for complete listing in bioc_VIEWS. If NA cross check is skipped and packages not built on any system will be missing from biocVIEWS |
meatPath |
character(1). File path location to the directory containing cloned repositories of Bioconductor packages. If manifestFile is used for cross checking and the meatPath is provided, entries from the DESCRIPTION file are manually entered into biocVIEWS information. If NA dummy values for minimal fields for landing page generation are included with ERROR. This attempts to fill in as much information as possible for packages that have failed to build. |
Seth Falcon
write_PACKAGES,
extractVignettes,
write_REPOSITORY,
write_VIEWS
This function returns a list of BiocView-class objects
corresponding to the subgraph of the views DAG induced by
topTerm. In short, this does the same thing as
getBiocViews, but limits the vocabulary to
topTerm and all of its decendents.
getBiocSubViews(reposUrl, vocab, topTerm, local = FALSE, htmlDir = "")getBiocSubViews(reposUrl, vocab, topTerm, local = FALSE, htmlDir = "")
reposUrl |
URL for a CRAN-style repository that hosts a
|
vocab |
A |
topTerm |
A string giving the name of the subview DAG. This view and all of its decendents will be included in the result. |
local |
logical indicating whether to assume a local package
repository. The default is |
htmlDir |
if the |
The root of the vocabulary DAG is implicitly included in the view creation process order to build views with a link back to the top. It is removed from the return list.
This function is tailored to generation of Bioconductor Task Views.
With the current vocabulary, it probably only makes sense to call it
with topView set to one of "Software",
"AnnotationData", or "ExperimentData". This is a hack
to allow the biocViews code to manage HTML views across more than one
repository.
A list of BiocView-class objects. The names of the list give
the name of the corresponding view.
Seth Falcon
data(biocViewsVocab) reposPath <- system.file("doc", package="biocViews") reposUrl <- paste("file://", reposPath, sep="") biocViews <- getBiocSubViews(reposUrl, biocViewsVocab, "Software") print(biocViews[1:2])data(biocViewsVocab) reposPath <- system.file("doc", package="biocViews") reposUrl <- paste("file://", reposPath, sep="") biocViews <- getBiocSubViews(reposUrl, biocViewsVocab, "Software") print(biocViews[1:2])
Given the URL to a CRAN-style package repository containing a
VIEWS file at the top-level and a graph-class object
representing a DAG of views, this function returns a list of
BiocView-class objects.
getBiocViews(reposUrl, vocab, defaultView, local = FALSE, htmlDir = "")getBiocViews(reposUrl, vocab, defaultView, local = FALSE, htmlDir = "")
reposUrl |
URL for a CRAN-style repository that hosts a
|
vocab |
A |
defaultView |
A string giving the term to use for packages that
do not list a term of their own via the |
local |
logical indicating whether to assume a local package
repository. The default is |
htmlDir |
if the |
A list of BiocView-class objects. The names of the list give
the name of the corresponding view.
Seth Falcon
data(biocViewsVocab) reposPath <- system.file("doc", package="biocViews") reposUrl <- paste("file://", reposPath, sep="") biocViews <- getBiocViews(reposUrl, biocViewsVocab, "NoViewProvided") print(biocViews[1:2])data(biocViewsVocab) reposPath <- system.file("doc", package="biocViews") reposUrl <- paste("file://", reposPath, sep="") biocViews <- getBiocViews(reposUrl, biocViewsVocab, "NoViewProvided") print(biocViews[1:2])
This function looks returns a list containing all the biocViews that are present on the Bioconductor website.
getCurrentbiocViews()getCurrentbiocViews()
It parses the dot file present inside the biocViews package.
It returns a named list with 3 components.
Software |
biocViews from the software branch |
ExperimentData |
biocViews from the ExperimentData branch |
AnnotationData |
biocViews from the AnnotationData branch |
Sonali Arora
ans <- getCurrentbiocViews() ## only the first 6 from each branch are shown here. lapply(ans, head)ans <- getCurrentbiocViews() ## only the first 6 from each branch are shown here. lapply(ans, head)
These functions visit two Bioconductor releases, identifying packages that are present in the ‘current’ repository and have NEWS since the base version of the same package in the ‘previous’ release. All NEWS is reported for packages only in the current repository.
getPackageNEWS(prevRepos="3.6", currRepos="3.7", repo=c("bioc", "data/experiment", "workflows"), srcdir = NULL) printNEWS(dbs, destfile, overwrite = FALSE, width = 68, output=c("md", "text"), relativeLink=FALSE, ...)getPackageNEWS(prevRepos="3.6", currRepos="3.7", repo=c("bioc", "data/experiment", "workflows"), srcdir = NULL) printNEWS(dbs, destfile, overwrite = FALSE, width = 68, output=c("md", "text"), relativeLink=FALSE, ...)
prevRepos |
|
currRepos |
|
repo |
|
srcdir |
Path to local checkout of package repositories, if NULL will try and use files on main builders |
dbs |
A list of |
destfile |
|
overwrite |
|
width |
|
output |
|
relativeLink |
Should links to packages be relative links on bioconductor.org website or include full url 'https//bioconductor.org'. default: FALSE is full url. |
... |
additional arguments, unused. |
A list of news_db files, as returned by utils::news,
for each package for which relevant NEWS is available.
Martin Morgan [email protected] and Lori Shepherd
These functions visit two Bioconductor releases branches, identifying differnt packages that are present in the ‘current’ repository from the ‘previous’ release. Utilizes the devel branch of bioconductor to retrieve description.
getPackageTitles(prevBranch="RELEASE_3_6", currBranch="devel", manifest=c("software.txt", "data-experiment.txt", "workflows.txt", "data-annotation.txt", "books.txt"), status = c("new", "removed")) getPackageDescriptions(pkgs, outfile, output=c("md", "text"), relativeLink=FALSE)getPackageTitles(prevBranch="RELEASE_3_6", currBranch="devel", manifest=c("software.txt", "data-experiment.txt", "workflows.txt", "data-annotation.txt", "books.txt"), status = c("new", "removed")) getPackageDescriptions(pkgs, outfile, output=c("md", "text"), relativeLink=FALSE)
prevBranch |
|
currBranch |
|
manifest |
|
status |
get new or removed package list comparing currBranch to prevBranch |
pkgs |
character() A list of packages to retrieve DESCRIPTION |
outfile |
|
output |
|
relativeLink |
Should links to packages be relative links on bioconductor.org website or include full url 'https//bioconductor.org'. default: FALSE is full url. |
A list of package titles.
Martin Morgan [email protected] and Lori Shepherd
## Not run: # At release time get a list of new or removed or deprecated packages # get new packages in release 3.7 that are not in 3.6 newSoft = getPackageTitles() # get removed packages from 3.6 rmSoft = getPackageTitles(currBranch="RELEASE_3_7", status="removed") # get depreacted package for 3.7 deprecatedSoft = setdiff(getPackageTitles(status="removed"), rmSoft) # repeated above for data-experiment packages newData = getPackageTitles(manifest="data-experiment.txt") rmData = getPackageTitles(currBranch="RELEASE_3_7", manifest="data-experiment.txt", status="removed") deprecatedData = setdiff(getPackageTitles(manifest="data-experiment.txt", status="removed"), rmData) ## End(Not run)## Not run: # At release time get a list of new or removed or deprecated packages # get new packages in release 3.7 that are not in 3.6 newSoft = getPackageTitles() # get removed packages from 3.6 rmSoft = getPackageTitles(currBranch="RELEASE_3_7", status="removed") # get depreacted package for 3.7 deprecatedSoft = setdiff(getPackageTitles(status="removed"), rmSoft) # repeated above for data-experiment packages newData = getPackageTitles(manifest="data-experiment.txt") rmData = getPackageTitles(currBranch="RELEASE_3_7", manifest="data-experiment.txt", status="removed") deprecatedData = setdiff(getPackageTitles(manifest="data-experiment.txt", status="removed"), rmData) ## End(Not run)
Given a repository URL, download and parse the VIEWS file.
getPacksAndViews(reposURL, vocab, defaultView, local=FALSE)getPacksAndViews(reposURL, vocab, defaultView, local=FALSE)
reposURL |
character vector giving the URL of a CRAN-style repository containing a VIEWS file at the top-level. |
vocab |
A |
defaultView |
A string giving the term to use for packages that
do not list a term of their own via the |
local |
logical indicating whether certain links should be
absolute (using |
A list with named elements:
views: Vector of view memberships. Names are package names.
pkgList: A list of PackageDetail-class objects.
Seth Falcon
Given a Directed Acyclic Graph (DAG) represented as a graphNEL
instance, return a character vector consisting of the specified
term and all of its descendants. That is, give the list of
terms for which a path exists starting at term.
getSubTerms(dag, term)getSubTerms(dag, term)
dag |
A |
term |
A string giving a term in the vocabulary (a node in |
A character vector of term names.
S. Falcon
data(biocViewsVocab) getSubTerms(biocViewsVocab, "Software")data(biocViewsVocab) getSubTerms(biocViewsVocab, "Software")
biocViews are "keywords" which are used to describe a given package. They are broadly divided into three categories, representing the type of packages present in the Bioconductor Project - Software, Annotation Data and Experiment Data. biocViews are supposed to come from only one of the three fields, but this function will check the list of biocViews and guess the package type based on how many biocViews came from each field.
guessPackageType(biocViews)guessPackageType(biocViews)
biocViews |
A character vector containing a list of biocViews. |
A character(1) of package type: either "Software", "ExperperimentData", or "AnnotationData".
Lori Shepherd
guessPackageType(c("clustering", "classification")) guessPackageType(c("Organism", "Homo Sapien"))guessPackageType(c("clustering", "classification")) guessPackageType(c("Organism", "Homo Sapien"))
This generic function should return an XMLNode instance
representing the specified object in HTML as a complete HTML document.
htmlDoc(object, ...)htmlDoc(object, ...)
object |
An object |
... |
Not currently used. |
An instance of XMLNode from the XML package.
Seth Falcon
This function returns a string containing an appropriate filename for storing the object's HTML representation.
htmlFilename(object, ...)htmlFilename(object, ...)
object |
An object. |
... |
Not currently used |
A character vector of length one containing the filename.
Seth Falcon
A virtual class for HTML serialization method dispatch.
A virtual Class: No objects may be created from it.
signature(object = "Htmlized"): Return the
html-ized representation of object as a complete HTML document.
Seth Falcon
This generic function should return an XMLNode instance
representing the specified object in HTML
htmlValue(object)htmlValue(object)
object |
An object |
An instance of XMLNode from the XML package.
Seth Falcon
Representation of R package metadata. Most slots correspond to fields in a package's DESCRIPTION file.
Objects can be created by calls of the form new("PackageDetail", ...).
Package:Object of class "character" see DESCRIPTION
Version:Object of class "character" see
DESCRIPTION
Title:Object of class "character" see
DESCRIPTION
Description:Object of class "character" see
DESCRIPTION
Author:Object of class "character" see DESCRIPTION
Maintainer:Object of class "character" see
DESCRIPTION
Depends:Object of class "character" see DESCRIPTION
Imports:Object of class "character" see DESCRIPTION
Suggests:Object of class "character" see DESCRIPTION
SystemRequirements:Object of class "character"
see DESCRIPTION
License:Object of class "character" see DESCRIPTION
URL:Object of class "character" see DESCRIPTION
biocViews:Object of class "character" see
DESCRIPTION
vignettes:Object of class "character" giving
paths to vignette pdf files in the repository
vignetteScripts:Object of class "character" giving
paths to vignette Stangled R files in the repository
vignetteTitles:Object of class "character" giving
the titles of the vignette files in the repository
source.ver:Object of class "character" version
string for the source package
win.binary.ver:Object of class "character"
version string for the 32-bit Windows binary package
mac.binary:Object of class "character"
version string for the macOS High Sierra binary package
mac.binary.mavericks.ver:Object of class "character"
version string for the OS X Mavericks binary package
mac.binary.el-capitan.ver:Object of class "character"
version string for the OS X El Capitan binary package
downloadStatsUrl:Object of class "character" An
optional URL for the download history statistics.
manuals:Object of class "character" giving paths to
reference manual pdf files in the repository
dependsOnMe:Object of class "character" giving
packages found in the repository that depend on this package
importsMe:Object of class "character" giving
packages found in the repository that imports this package
suggestsMe:Object of class "character" giving
packages found in the repository that suggest this package
functionIndex:Object of class "character" Not
used. Intended to hold function index data.
reposFullUrl:Object of class "character" The URL
for the full URL of the root of the repository.
reposRoot:Object of class "character" The URL
for the root of the repository.
viewRoot:Object of class "character" The URL
for the view of the repository.
devHistoryUrl:Object of class "character" The URL
for the development changelog.
Class "Htmlized", directly.
signature(object = "PackageDetail"): Return an
XMLNode instance containg a complete HTML document
representation of the package.
signature(object = "PackageDetail"):
Return a filename appropriate for the HTML document representation.
signature(object = "PackageDetail"): Return
XMLNode instance containing an HTML representation of the package.
pdAuthorMaintainerInfo-class
pdVignetteInfo-class
pdDownloadInfo-class
pdDetailsInfo-class
pdDescriptionInfo-class
pdVigsAndDownloads-class
Dummy classes for HTML generation. Each dummy class is a simple
extension (it does not add any slots). The purpose of each dummy
class is to allow for method dispatch to generate HTML via the
htmlValue method.
You can convert convert a PackageDetail instance to one of the
dummy classes like this: descInfo <- as(pdObj,
"pdDescriptionInfo")
Seth Falcon
pd <- new("PackageDetail", Package="MyFancyPackage", Version="1.2.3", Title="A Fancy Package", Description="This package does fancy things", Author="A. Coder", Maintainer="A. Coder <[email protected]>", Depends="methods", Imports="ASimplePackage", Suggests="MyDataPackage", biocViews="Infrastructure", vignettes="vignettes/MyFancyPackage/inst/doc/MFP1.pdf,\nvignettes/MyFancyPackage/inst/doc/MFP2.pdf", vignetteScripts="vignettes/MyFancyPackage/inst/doc/MFP1.R\nvignettes/MyFancyPackage/inst/doc/MFP2.R", vignetteTitles="MFP1 Document,\nMFP2 Document", source.ver="src/contrib/MyFancyPackage_1.2.3.tar.gz", win.binary.ver="bin/windows/contrib/4.0/MyFancyPackage_1.2.2.zip", mac.binary.ver="bin/macosx/contrib/4.0/MyFancyPackage_1.2.3.tgz", dependsOnMe=c("PackageThatExposesMe"), importsMe=c("AnEvenFancierPackage","AMuchFancierPackage"), suggestsMe="PackageThatUsesMeInVignette", reposRoot="http://foo.bar.org") html <- htmlValue(pd) pdpd <- new("PackageDetail", Package="MyFancyPackage", Version="1.2.3", Title="A Fancy Package", Description="This package does fancy things", Author="A. Coder", Maintainer="A. Coder <[email protected]>", Depends="methods", Imports="ASimplePackage", Suggests="MyDataPackage", biocViews="Infrastructure", vignettes="vignettes/MyFancyPackage/inst/doc/MFP1.pdf,\nvignettes/MyFancyPackage/inst/doc/MFP2.pdf", vignetteScripts="vignettes/MyFancyPackage/inst/doc/MFP1.R\nvignettes/MyFancyPackage/inst/doc/MFP2.R", vignetteTitles="MFP1 Document,\nMFP2 Document", source.ver="src/contrib/MyFancyPackage_1.2.3.tar.gz", win.binary.ver="bin/windows/contrib/4.0/MyFancyPackage_1.2.2.zip", mac.binary.ver="bin/macosx/contrib/4.0/MyFancyPackage_1.2.3.tgz", dependsOnMe=c("PackageThatExposesMe"), importsMe=c("AnEvenFancierPackage","AMuchFancierPackage"), suggestsMe="PackageThatUsesMeInVignette", reposRoot="http://foo.bar.org") html <- htmlValue(pd) pd
Packages being added to the Bioconductor Project require biocViews in their DESCRIPTION file.(Note that the field name "biocViews" is case-sensitive and must begin with a lower-case 'b'.)biocViews are "keywords" which are used to describe a given package. They are broadly divided into three categories, representing the type of packages present in the Bioconductor Project - Software, Annotation Data and Experiment Data.
recommendBiocViews(pkgdir, branch)recommendBiocViews(pkgdir, branch)
pkgdir |
The path of the package Directory. |
branch |
The branch which your package will belong to. It can be either 'Software', 'AnnotationData' or 'ExperimentData'. |
This function parses the package directory provided by the user to recommend biocViews to the user. The output is a suggested list - the user of this function is expected to go through this list and find which biocViews best describe his or her package. It uses the following strategies.
It parses the "Description", "Title", "Package" of the DESCRIPTION page to find biocViews.
It looks up the biocViews of the packages in the "Depends" field of the given package to recommend biocViews
It parses the text from the man pages and the vignettes to suggest biocViews.
Please note the following:
Do not make up your own biocViews.
Double check the spelling and case of the biocViews added.
Please add biocViews only from the appropriate branch. eg: Software packages should have only Software biocViews.
A list is returned with 3 characters - current , recommended and remove.
"current" contains the biocViews from the package's DESCRIPTION file.
"recommended" are the recommended biocViews - This is a suggested list which the user can add in addition to "current" biocViews - the user is expected to go through this list and find which biocViews best describe their package.
"remove" are those biocViews which are inconsistent with the Bioconductor biocViews. (Hint - check for spelling, cases and plural)
Sonali Arora.
biocViews are "keywords" which are used to describe a given package. They are broadly divided into three categories, representing the type of packages present in the Bioconductor Project - Software, Annotation Data and Experiment Data. One can find packages which are tagged with biocViews using this function.
recommendPackages(biocViews, use.release=TRUE, intersect.views=TRUE)recommendPackages(biocViews, use.release=TRUE, intersect.views=TRUE)
biocViews |
A character vector containing a list of biocViews. Currently only biocViews from the software branch are supported. |
use.release |
A logical character indicating if you want packages recommended from the release branch of Biocondutor. |
intersect.views |
A logical character indicating if you want packages which are tagged with all the input biocViews or packages tagged with any one or more of the biocViews. |
A character vector containing a list of packages. If multiple biocViews are given as input, the result returns packages tagged with all or atleast one of the input biocViews.
Sonali Arora.
recommendPackages(c("clustering", "classification"))recommendPackages(c("clustering", "classification"))
Representation of R package repository index
Objects can be created by calls of the form new("RepositoryDetail", ...).
Title:Object of class "character" giving the
title for the repository.
reposRoot:Object of class "character" giving the root
URL of the repository
homeUrl:Object of class "character" ?
htmlDir:Object of class "character" ?
packageList:Object of class "list" consisting
of objects of class PackageDetail-class
Class "Htmlized", directly.
signature(object = "RepositoryDetail"): ...
signature(object = "RepositoryDetail"): ...
signature(object = "RepositoryDetail"): ...
Seth Falcon
Ensures that a package has biocViews and that they are valid. Function is designed to be called from the unit tests of another package.
validate_bioc_views(pkg)validate_bioc_views(pkg)
pkg |
|
invisible(NULL) if tests pass.
Dan Tenenbaum [email protected]
validate_bioc_views("biocViews")validate_bioc_views("biocViews")
This function writes a REPOSITORY file at the top-level of a
CRAN-style repository. This file is DCF formatted and describes the
location of packages available in the repository. Here is an example
for a repository containing source packages, and Windows and Mac binary
packages:
source: src/contrib
win.binary: bin/windows/contrib/4.0
mac.binary: bin/macosx/contrib/4.0
provides: source, win.binary, mac.binary
write_REPOSITORY(reposRootPath, contribPaths)write_REPOSITORY(reposRootPath, contribPaths)
reposRootPath |
character vector containing the path to the CRAN-style repository root directory. |
contribPaths |
A named character vector. Valid names are
|
Seth Falcon
write_PACKAGES,
extractVignettes,
genReposControlFiles,
write_VIEWS
Writes a DCF formatted file, SYMBOLS, containing the symbols exported by each package in a directory containg R package source directories.
write_SYMBOLS(dir, verbose = FALSE, source.dirs=FALSE)write_SYMBOLS(dir, verbose = FALSE, source.dirs=FALSE)
dir |
The root of a CRAN-style package repository containing
source packages. When |
verbose |
Logical. When |
source.dirs |
Logical. When |
Returns NULL. Called for the side-effect of creating a
SYMBOLS file in dir.
S. Falcon
This function writes a VIEWS file to the top-level of a
CRAN-style package repository. The VIEWS file is in DCF format
and describes all packages found in the repository.
The VIEWS file contains the complete DESCRIPTION file
for each source package in the repository. In addition, metadata for
available binary packages and vignettes is centralized here.
write_VIEWS(reposRootPath, fields = NULL, verbose = FALSE, vignette.dir = "vignettes", manifestFile = NA, meatPath = NA)write_VIEWS(reposRootPath, fields = NULL, verbose = FALSE, vignette.dir = "vignettes", manifestFile = NA, meatPath = NA)
reposRootPath |
character vector containing the path to the CRAN-style repository root directory. |
fields |
Any additional fields to include. You shouldn't need this, but if you have added fields to the DESCRIPTION files of the packages in the repository, you may want it. |
verbose |
logical, if |
vignette.dir |
character specifying where to look for vignettes. |
manifestFile |
character(1). File path location to Bioconductor formatted manifest file that lists all current packages. This file will be used in the write_VIEWS function to cross check successfully built packages with all expected packages. Packages that have not built will be given dummy entry for complete listing in bioc_VIEWS. If NA cross check is skipped and packages not built on any system will be missing from biocVIEWS |
meatPath |
character(1). File path location to the directory containing cloned repositories of Bioconductor packages. If manifestFile is used for cross checking and the meatPath is provided, entries from the DESCRIPTION file are manually entered into biocVIEWS information. If NA dummy values for minimal fields for landing page generation are included with ERROR. This attempts to fill in as much information as possible for packages that have failed to build. |
This function uses a private function from the tools package:
tools:::.build_repository_package_db.
Seth Falcon
write_PACKAGES,
extractVignettes,
genReposControlFiles,
write_REPOSITORY
This function serializes a list of BiocView-class
objects to a series of HTML files.
writeBiocViews(bvList, dir, backgroundColor="transparent")writeBiocViews(bvList, dir, backgroundColor="transparent")
bvList |
A list of |
dir |
A character vector giving the directory where the HTML files will be written. |
backgroundColor |
A character vector giving the background color for the body in the CSS file. |
Seth Falcon
getBiocViews,
genReposControlFiles,
write_VIEWS
Given a DOM tree from the XML package and a filename, write the DOM to disk creating an HTML file.
writeHtmlDoc(html, file)writeHtmlDoc(html, file)
html |
A DOM object from the XML package |
file |
A string giving the filename |
S. Falcon
This function creates package "homepages" that describe the package and provide links to download package artifacts in the repository.
writePackageDetailHtml(pkgList, htmlDir = "html", backgroundColor="transparent")writePackageDetailHtml(pkgList, htmlDir = "html", backgroundColor="transparent")
pkgList |
A list of |
htmlDir |
The files will be written to this directory. |
backgroundColor |
A character vector giving the background color for the body in the CSS file. |
Seth Falcon
This function generates an HTML file for each package in a repository
and generates an index.html file that provides an alphabetized
listing of the packages.
writeRepositoryHtml(reposRoot, title, reposUrl = "..", viewUrl = "../..", reposFullUrl=reposUrl, downloadStatsUrl="", devHistoryUrl="", link.rel = TRUE, backgroundColor="transparent")writeRepositoryHtml(reposRoot, title, reposUrl = "..", viewUrl = "../..", reposFullUrl=reposUrl, downloadStatsUrl="", devHistoryUrl="", link.rel = TRUE, backgroundColor="transparent")
reposRoot |
string specifying the path to the root of the CRAN-style package repository. |
title |
string giving the title for the repository |
reposUrl |
string giving the prefix for URL in links
generated on the package description pages. The default is
|
viewUrl |
string giving the prefix for the URL in links to the view pages. The biocViews terms will be linked to views summary pages with this prefix. |
reposFullUrl |
string giving the full prefix for URL in links
generated on the package description pages. The default is
|
downloadStatsUrl |
string giving the prefix for the URL in links to the download history statistics pages. |
devHistoryUrl |
string giving the prefix for the URL in links to the development changelog. |
link.rel |
logical indicating whether the index page should
generate relative URL links. The default is |
backgroundColor |
A character vector giving the background color for the body in the CSS file. |
Seth Falcon
Ensures that .R files from vignette code chunks are written out.
writeRFilesFromVignettes(reposRoot, reposUrl="..", viewUrl="../..", reposFullUrl=reposUrl, downloadStatsUrl="", devHistoryUrl="")writeRFilesFromVignettes(reposRoot, reposUrl="..", viewUrl="../..", reposFullUrl=reposUrl, downloadStatsUrl="", devHistoryUrl="")
reposRoot |
Root directory of a CRAN-style repository |
reposUrl |
URL of repository |
viewUrl |
url of VIEWS file |
reposFullUrl |
Full URL of VIEWS file |
downloadStatsUrl |
URL to download stats page |
devHistoryUrl |
Dev history URL |
Given a directory and a vocabulary represented as a graphNEL
containing a DAG of terms, write the top-level term to disk as HTML.
This assumes your vocabulary has a single term with no parents.
writeTopLevelView(dir, vocab)writeTopLevelView(dir, vocab)
dir |
A string giving a directory in which to write the HTML file |
vocab |
A |
S. Falcon