About Lectures Research Software Blog
Musical Site
MySpace
Facebook

Moods Blog

Dojo Shin Kaï

RSS Feed
Thank you!

XHTML 1.0 conformant
CSS 2.0 conformant
Didier Verna's scientific blog: Lisp, Emacs, LaTeX and random stuff.

Tag - Common Lisp

Entries feed - Comments feed

Tuesday, May 10 2022

Declt 4.0 beta 1 "William Riker" is released

Today, after two years and a half of (irregular) development, and 465 commits, I have released the first beta version of Declt 4.0, my reference manual generator for Common Lisp libraries.

I seldom release official beta versions of software, let alone make announcements about them, but this one deserves an exception. Since version 3, Declt has been undergoing a massive, SOC-oriented, overhaul. The intent is to reimplement it into a clean 3-stages pipeline. Stage one, the "assessment" stage is in charge of gathering information. Stage two, the "assembly" stage, creates a documentation template specifying how the information is to be organized. Finally, Stage three, the "typesetting" stage, renders the template in a specific output format. Of course, the purpose is to eventually be able to support different kinds of templates and output formats.

Declt 4.0b1 marks the achievement of Stage one, which is now complete. It also contains a lot of new (and long awaited) features and improvements. The user manual has been updated to provide an overview of the new architecture, along with some details about Stage one of the pipeline. From now on, I'm also going to release new versions of Quickref to closely follow the evolution of Declt.

Apart from the aforementioned architecture overhaul, which really boils down to internals shaking and shouldn't affect the end-user much, a lot of new features and improvements are also provided in this release. They are listed below.

Backward Incompatible Changes

For the benefit of Separation of Concerns, but also for other reasons, a number of keyword arguments to the declt function have been renamed. :hyperlinks becomes :locations, :version becomes :library-version, :texi-directory becomes :output-directory, and :texi-name becomes :file-name.

New Features

Default / Standard Values

Declt now has the ability to advertise or hide properties that get default / standard values (e.g. standard method combination, :instance slot allocation, etc.).

Support for Aliases

Declt now recognizes and advertises "aliases", that is, funcoids which have their fdefinition, macro, or compiler macro function set manually to another, original, definition.

Support for New Programmatic Entities

Typed structures and setf compiler macros are now properly detected and documented with all the specific information.

Proper Support for Uninterened Symbols

Such symbols may be encountered on several occasions such as slot names (trivialib, for example, defines structures with uninterned slot names in order to prevent access to them). Definitions named with uninterned symbols are considered private, and denoted by the empty set package (∅).

SBCL 2.1.2 Required

The following enhancements depend on it:

  • short form setf expanders now get correct source information,
  • method combinations lambda-lists are now documented.

Domestic vs. Foreign Definitions

The concept of domestic (as opposed to foreign) definition has been extended to include those created in one of the sources files of the library being documented, even if the symbol naming the definition is from a foreign package. If the source file is unknown, then the symbol's package must be domestic. The general intent is to consider language extensions as domestic, hence, always documented. For example, a new method on a standard generic function (say, initialize-instance) will now always appear in the documentation.

This refinement is accompanied by a new option allowing to include foreign definitions in the generated documentation. Of course, only foreign definitions somehow related to the library being documented will appear in the documentation. Also, those definitions will in general be partial: only the parts relevant to the library being documented will appear. Continuing with the initialize-instance example, new methods normally appear as toplevel, standalone definitions in the documentation (and their parent generic function is simply mentioned). If foreign definitions are included however, there will be a toplevel entry for the generic function, and all methods defined in the library will be documented there (truly foreign methods won't appear at all).

Introspection Heuristic

Until now, there was a single heuristic for finding domestic definitions, which was to start from domestic packages and scan as many connections between symbols as possible. That heuristic is reasonably fast, but may occasionally miss some domestic definitions. Declt now has the ability to scan all symbols in the Lisp image instead of only the ones from domestic packages. This new option ensures that all domestic definitions are found, at the expense of a much greater computation time.

Supported Licenses

The Microsoft Public License has been added.

Info Installation Category

In other words, the value of Texinfo's @direntry command is now customizable (instead to being hardwired to "Common Lisp").

Improvements

Documentation Thinning

Packages reference lists do not point to methods directly anymore (as they can be reached via the generic function's reference). Also, only slots for which the parent classoid is named from another package are now referenced.

Files reference lists do not point to slots anymore (as they can be reached via the parent classoid's reference). Also, only methods for which the parent generic function is defined in another file are now referenced.

The readability of long references has been improved. In particular, they don't advertise the type of the referenced definitions anymore, when there is no ambiguity.

Slot documentation now advertises the slot name's package only when different from that of the parent classoid.

Non standalone method documentation now advertises the source file only when different from that of the parent generic function.

The rendering of EQL specializers has been inmproved.

The documentation of setf / writer methods doesn't render the "new value" argument / specializer anymore.

Merging

Generic definitions containing only reader / writer methods are upgraded to specific reader / writer categories, and definition merging is now only attempted on those.

Lambda Lists

Uninformative parts of lambda lists are now filtered out. This includes &whole, &environment, &aux variables, along with options / keyword variables and default values.

Method specializers in lambda lists now link back to their respective class definitions.

Method Combinations

The method combination discovery scheme has been upgraded to benefit from SBCL 1.4.8's enhancements (themselves following my ELS 2018 paper). The old code didn't break, but prevented unused method combinations from being detected.

Bug Fixes and Workarounds

ASDF

Better handling of missing / unloaded components or dependencies (this can happen for instance with feature-dependent conditional inclusion).

Cope with the lack of specification of the license information in ASDF systems by coercing to a string.

Fix several cases of system files documentation duplication. Declt automatically documents .asd files as special cases of Lisp files. However, some systems have the bad (IMHO) habit of mentioning them explicitly as components (e.g. static files). When this happens, Declt silently discards that definition and keeps its own (at the expense of having a slightly incorrect system documentation).

Anchoring

After years of vain attempts at providing human-readable yet unique anchor names (which was only really useful for Info, BTW), I finally got rid of my last bit of idealism. Anchor names now use numerical definition UIDs, and since Texinfo allows me to expand references with some more information than just the anchor, it's good enough. Besides, it fixes the last remaining rare cases exhibiting the fact that it's just impossible to have anchors that are both human-readable and unique.

Setf Expanders

Fix the computation of short form setf epxander lambda lists (which didn't correctly handle the presence of optional or rest arguments before.

Handle the potential unavailability of a setf expander's update function or short form operator. Document it if applicable. Also signal a warning when the expander is domestic.

Monday, March 16 2020

TFM 1.1 "Carolingan Miniscules" is released

I'm happy to announce the release of TFM version 1.1 "Carolingan Miniscules". TFM is a TeX Font Metrics parser library for Common Lisp.

This release provides support for font scaling and freezing. Scaling is done, as in TeX, by overriding the font's design size, all other dimensions remaining relative to it. Freezing, on the other hand, converts all relative dimensions into absolute ones, potentially saving a lot of run-time arithmetic computation.

Get it at the usual place.

Thursday, November 21 2019

Quickref 3.0 "The Alchemist" is released

Following up with the latest release of Declt, I'm also happy to announce the release of Quickref 3.0 "The Alchemist", our reference manuals aggregator for Quicklisp libraries. The official website has been updated yesterday with both the latest Quicklisp version, and the new Declt, resulting in the documentation of 1792 Common Lisp libraries.

Tuesday, November 19 2019

Declt 3.0 "Montgomery Scott" is released

I'm pleased to announce the release of Declt 3.0 "Montgomery Scott", my Texinfo reference manual generator for Common Lisp libraries.

This is a new major release of the library, although the changes may not be so visible from the outside. The main concern in this release has been to increase the robustness of the output, from three different angles.

  1. Many places from where Declt attempts to extract meaningful information are underspecified (ASDF system slots notably).
  2. The pretty-printing of Lisp items can be difficult, given the liberalism and extensibility of the language.
  3. Finally, several restrictions in the syntax of Texinfo itself (anchor names notably) get in the way.

These issues were described in a paper presented at the TeX Users Group conference, this summer in Palo Alto. This release goes a long way toward fixing them.

The new generated reference manuals look mostly the same as before, but some important things have changed under the hood, notably the names of hyperlinks and cross-references (backward-incompatible, hence the increase in the major version number). The output is now also much more robust with respect to the final output format: the generation of HTML, DVI, Postscript, PDF, and even plain Info should work fine and has been tested on all Quicklisp libraries (in fact, Quickref will also be upgraded in the near future to provide all those formats at once).

Those improvements do come at a cost. Unicode is now required (even for Info readers). To be honest, many Lisp libraries already had that implicit requirement. Also, this release depends on improvements and bug fixes only available in Texinfo 6.7, now a requirement as well. I have to thank Gavin Smith, from the Texinfo team, for his collaboration.

Apart from that, this release also has a number of bug fixes.

  • Some method specializers were not handled properly.
  • The manuals were missing documentation for some non-Lisp source files.
  • There were some glitches in the pretty-printing of unreadable objects.

Get it at the usual place, and enjoy!

Monday, April 8 2019

Quickref 2.0 "Be Quick or Be Dead" is released

Surfing on the energizing wave of ELS 2019, the 12 European Lisp Symposium, I'm happy to announce the release of Quickref 2.0, codename "Be Quick or Be Dead".

The major improvement in this release, justifying an increment of the major version number (and the very appropriate codename), is the introduction of parallel algorithms for building the documentation. I presented this work last week in Genova so I won't go into the gory details here, but for the brave and impatient, let me just say that using the parallel implementation is just a matter of calling the BUILD function with :parallel t :declt-threads x :makeinfo-threads y (adjust x and y as you see fit, depending on your architecture).

The second featured improvement is the introduction of an author index, in addition to the original one. The author index is still a bit shaky, mostly due to technical problems (calling asdf:find-system almost two thousand times simply doesn't work) and also to the very creative use that some library authors have of the ASDF author and maintainer slots in the system descriptions. It does, however, a quite decent job for the majority of the authors and their libraries'reference manuals.

Finally, the repository now has a fully functional continuous integration infrastructure, which means that there shouldn't be anymore lags between new Quicklisp (or Quickref) releases and new versions of the documentation website.

Thanks to Antoine Hacquard, Antoine Martin, and Erik Huelsmann for their contribution to this release! A lot of new features are already in the pipe. Currently documenting 1720 libraries, and counting...

Wednesday, December 13 2017

Announcing Quickref: a global documentation project for Common Lisp

Today, we deployed the first version of Quickref, a new global documentation project for Common Lisp.

The purpose of Quickref is to provide a centralized collection of reference manuals for the whole Quicklisp world. This means around 1500 libraries, for a total of around 3000 ASDF systems. The reference manuals are generated by Declt, which is probably the most complete documentation system for Common Lisp currently available, and delivered in HTML (PDF versions could easily be made available as well).

A lot of things can still be improved, but I'm pretty satisfied with the result so far. 3000 ASDF systems is a hell of a test suite for Declt, and I'm happy to report that it passes on practically all of them. Only a couple of issues remain, not even due to Declt itself, and less than a dozen or so libraries still pose problems (mostly technical difficulties due to foreign dependencies).

Quickref was made by Antoine Martin, as part of an internship with me. Many thanks to him! We still have some cleanup and packaging to do, but we expect to open-source the infrastructure soon. I also want to thank Mark Evenson, Erik Huelsmann and the Common Lisp Foundation for hosting the project on common-lisp.net (it was only natural)!

Finally, let me restate this again (and again): reference manuals are not user manuals. They are... reference manuals. Although automatically generated, there are some things you can do, as a library author, to improve the output (this is an area of Declt which I intend to work on in the future). Please refer to the Declt user manual (notably section 3.2 Coding Style) for more information.

Friday, October 27 2017

Standard IO syntax and the Robustness Principle

Here is a flagrant illustration of the robustness principle, or rather, of a failure to honor it.

I was investigating a bug in Declt where some floating point numbers were printed with exponent markers (e.g. 0.5f0 instead of just 0.5) in the Texinfo file, which broke the parsing of the file by Perl.

Eventually, I found out a double infringement of the robustness principle. First of all, Declt failed to comply with part 1 of the robustness principle: be lenient with the others. The Texinfo file generation routine should have been wrapped into a call to WITH-STANDARD-IO-SYNTAX and it wasn't. Always do that to be on the safe side. Lesson learnt.

This failure on my part, however, had the interesting consequence of exhibiting what I consider a serious infringement of part 2 of the robustness principle: be strict with yourself. It would have remained unnocited otherwise. The culprit here is not Declt. This time, it's the common-lisp-stat library. Problem: the simple fact of loading this library globally changes the value of *READ-DEFAULT-FLOAT-FORMAT* from SINGLE-FLOAT (the default) to DOUBLE-FLOAT. This is bad, and it can break your code in all sorts of nasty ways.

Explanation

*READ-DEFAULT-FLOAT-FORMAT* tells the reader how to read floats when no exponent marker is provided. By default, 0.5 will be read as a SINGLE-FLOAT. But this variable also influences the printer (out of a concern for printing readably I guess): when printing a float of a different format than the current default, then the appropriate exponent marker is also printed. So here is precisely what happened. Declt had been compiled with some floats (e.g. 0.5) read as SINGLE-FLOATs. Later on, those floats were supposed to be printed aesthetically as such. But right after loading common-lisp-stat, the default format changed to DOUBLE-FLOAT and all of a sudden 0.5 started to be printed as 0.5f0.

Consequences

This is bad enough already, but consider that messing with the standard IO syntax globally like this can break your code in all other sorts of even nastier ways. Imagine for instance that common-lisp-stat had been loaded before Declt, and Declt needed to be recompiled. All of a sudden, Declt would be using double floats and the bug would be gone. That is, until the next start of the REPL, after which all floats would be printed like 0.5d0!

So granted, my code wasn't robust enough. But please, don't mess with standard IO syntax globally!

Monday, October 16 2017

Declt 2.3 "Robert April" is out

I'm happy to announce the release of Declt 2.3. Declt is my reference manual generator for Common Lisp libraries.

The improvements and bug fixes in the last two releases are the result of running Declt against the whole Quicklisp world (around 3000 ASDF systems for 1500 libraries). See this post for more information.

New in this release:

  • Advertise file extensions in references.
  • Advertise the type of foreign definitions.
  • More robust display and indexing of, and with, lambda-lists.
  • Use UTF8 special characters to denote invisble ones.
  • More robust support for Texinfo brace escaping.
  • Handle modules sharing the same location.
  • Ensure output is done with standard IO syntax.
  • Fix potential duplication of some (non-lisp) files and document all static files.
  • Fix potential duplication of packages documentation.

From the 2.2 "Christopher Pike" release (not previously advertised):

  • Require a UTF-8 environment.
  • Understand ASDF's notion of inferred system, and also be more protective against ASDF extensions.
  • Support for improper lambda lists (e.g. destructuring ones).
  • Improve contact defaulting code.
  • Update support for SBCL's setf expanders introspection.
  • Accept ASDF system designators.
  • Various bug fixes in the areas of method combinations, accessor definition merging and setf expanders.

Find it at the usual place...

Tuesday, February 28 2017

Declt 2.1 "Jonathan Archer" is out

I'm happy to announce the release of Declt 2.1 "Jonathan Archer".

New in this release:

  • Handle recent change in SBCL's SB-INT:INFO API.
  • Handle list of contacts (strings) in ASDF systems author and maintainer slots.
  • Some backward-incompatible changes in the keyword arguments to the DECLT function.
  • More hyperlinks between systems and source files.
  • More robust system's packages collection (no more code walking).
  • More robust handling of unavailable ASDF components.
  • More robust naming of cross-references.

Find it at the usual place...

Wednesday, November 4 2015

Declt 2.0.1 "Benjamin Sisko" is out

Declt 2.0.1 "Benjamin Sisko" is out. This is a bugfix release with one internal change (a special variable was not following the earmuffs convention) and one actual bugfix (the same Texinfo anchor was generated for symbols with the same name but in different packages).

Find it at the usual place...

Monday, October 12 2015

ASDF-FLV 2.0

I've just released version 2.0 of ASDF-FLV, my ASDF extension for supporting file-local variables (ala *PACKAGE*). The code hasn't changed, but as for my other libraries, the system and package names are now prefixed with net.didierverna. ASDF-FLV is also available on GitHub now.

The reason I'm doing this now is that at least two of my other libraries are going to use it in a mandatory way, either directly or indirectly (and in turn, that's because no implementation has bothered to implement CDR #9 yet ;-).

Wednesday, August 5 2015

The Return of Segfaults

I was watching the discussion between Gilad Bracha and Matthias Felleisen on gradual typing this afternoon (it's available on YouTube). This was the last event at the STOP workshop, part of ECOOP 2015 in Prague. I couldn't attend it because I was somewhere else (Curry On) at the time. The discussion is interesting, but if you go all the way, in the last 10 minutes or so, you will notice that Matthias seems to be completely obsessed with what he calls the "Return of the SegFaults".

Basically, the point is the following. Mathias dislikes the optional types in Common Lisp because it's opening the door to unsafety. Initially, any dynamic language has a sound type system (all type errors are caught; at run-time, yes, but they are caught). As soon as you introduce an optional type system ala Lisp, the compiler implementors are "pushed" to use the provided type information for optimization, hence weakening the type system and breaking soundness. It's the "return of segfauts".

Of course, I agree with that, at least on the principle. Yes, Common Lisp's weak, optional type system is an antiquated thing. However, it seems to me that Matthias is forgetting two important things on the practical level:

  1. by default in many implementations that I know, if not all of them, introducing type declarations doesn't actually break the soundness of the type system but leads to even more type checking. For example, (defun plus (a b) (+ a b)) works on every possible numerical value, but add (declare (type fixnum a b)) in there and it will suddenly stop working on anything else but integers. It's only if you require the compiler to optimize for speed at the expense of safety that you effectively weaken your type system.
  2. In practice, the risk of re-introducing segfaults in your application may be mitigated by the interactive aspect of the development (TDD made easier, simultaneous write / compile / run / test / debug phases etc.).

So my impression is that Matthias is largely exaggerating the problem, but I'm not really qualified to tell. That's why I would like to know from you guys, working with Lisp in the industry, writing large applications, lots of type annotations, and (declaim (optimize (speed 3) (safety 0) (debug 0)) (EDIT: that's exaggerated of course, I really mean breaking safety for performance reasons): how much of a problem the "return of the segfaults" really is in practice ?

As a side note, this reminds me of the dynamic vs. lexical scoping debate. Many people were and still are strongly opinionated against dynamic scoping by default. Of course, I too, at least in principle. But how dangerous dynamic scoping really is in practice (EDIT: I'm not talking about expressiveness, e.g. closures, here. Only unsafety.)? What I can tell you is that in the 15 years I was actively maintaining the XEmacs codebase, I may have run into name clashes due to dynamic scoping... twice.

Monday, July 13 2015

Declt 2.0 is out -- IMPORTANT

Declt 2.0 "Kathryn Janeway" is out. This release doesn't contain any change in functionality, yet deserves a major version upgrade since it contains 3 important changes: an infrastructure revamp (along the lines of what Clon endured not so long ago), a license switch from the GNU GPL to a BSD one, and finally a system / package name change. The prefix is now net.didierverna instead of com.dvlsoft. Do I need to apologize for this again? :-)

Find it at the usual place...

Wednesday, January 14 2015

Insider Threat Detection at Haystax

Reported to me by Craig Norvell: Haystax appears to have some interesting Lisp projects in the works. Combining Lisp, Prolog, RDF (AllegroGraph), and a BBN for insider threat detection solutions.

From their website:

Haystax's predictive models are continuously updated based on the prevailing threat environment making it highly suitable for both detection and continuous evaluation of threats. These unique models go beyond traditional web and business intelligence to enable organizations to achieve contextual real-time situational awareness by fusing all operationally relevant information - private, public, video and live feeds - into consolidated views to show patterns and identify threats that are usually buried in too much noise or not placed in proper context.

Here is a Technical paper, and the STIDS Conference website. If you are interested, it looks like you can also attend a webcast on January 21st.

Saturday, August 24 2013

Declt 1.0 is out

After 15 betas, I'm happy enough with the current state of Declt to finally make a 1.0 release.

A lot of things have changed since the previous version, sometimes in a backward-incompatible way. Many more items are documented (including, as you have recently seen, method combinations). In addition to the reference manual, generated by Declt itself, there is now a real user manual.

Declt is still SBCL-only, requires ASDF 3 and Texinfo 4 but generates code that is compatible with Texinfo 5. Also, beware, I've deleted the old repository and moved the project to GitHub. Below is a more precise description of what Declt currently does.

Declt (pronounce "dec'let") is a reference manual generator for Common Lisp libraries. It works by loading an ASDF system and introspecting its contents. The generated documentation contains the description for the system itself and its components (modules and files), the packages defined in that system and the definitions found in those packages.

Exported and internal definitions are listed separately. This allows the reader to have a quick view on the library's public API. Within each section, definitions are sorted lexicographically.

In addition to ASDF system components and packages, Declt documents the following definitions: constants, special variables, symbol macros, macros, setf expanders, compiler macros, functions (including setf ones), generic functions and methods (including setf ones), method combinations, conditions, structures, classes and types.

The generated documentation includes every possible bit of information that introspecting can provide: documentation strings, lambda lists (including qualifiers and specializers where appropriate), slots (including type, allocation and initialization arguments), definition source file etc.

Every documented item provides a full set of cross-references to related items: ASDF component dependencies, parents and children, classes direct methods, super and subclasses, slot readers and writers, setf expanders access and update functions etc.

Finally, Declt produces exhaustive and multiple-entry indexes for every documented item.

Reference manuals are generated in Texinfo format (compatible, but not requiring Texinfo 5). From there it is possible to produce readable / printable output in info, HTML, PDF, DVI and PostScript with tools such as makeinfo, texi2dvi or texi2pdf.

The Declt reference manual is the primary example of documentation generated by Declt itself.

Friday, August 16 2013

Lisp Corner Cases: Method Combinations

In the process of writing Declt, I had to deepen my knowledge of some Lisp corner cases, notably in the area of introspection. As you know, Lisp has in fact more than 2 namespaces. Sometimes, introspecting a potential symbol definition in one namespace is trivial. COMPILER-MACRO-FUNCTION is one example. The "functional" namespace is heterogeneous but you can still make your way out of macros, regular or generic functions very easily. In the same vein, distinguishing constants, special variables and symbol macros in the "variables" namespace is more complicated because there is no standard way to access that information, but with the help of some vendor-specific machinery (e.g. SB-INT:INFO), it's still doable.

At some point, I tackled the case of method combinations, and to my greatest surprise, found out that introspecting a method combination by name is not simple. In fact, it's not even doable. The reason is that method combinations don't have a namespace proper. Let me explain why.

You define a new method combination of some NAME with DEFINE-METHOD-COMBINATION, and you use it with the :METHOD-COMBINATION option to DEFGENERIC. But how do you introspect a NAME for a potential method combination definition? Well, you can't do exactly that. First, there is no standard way to do so. Next, let's look at what the MOP provides. One would expect something along the lines of FIND-CLASS...

There is indeed something called FIND-METHOD-COMBINATION, but either I'm missing something, or it is really, and I mean really badly designed. The arguments to this function are: a generic function meta-object, a method combination name and some method combination options. So if this function is supposed to be the equivalent of FIND-CLASS for method combinations, what in the hell are the 1st and 3rd arguments for? In fact, it's not supposed to do what its name suggests. According to the AMOP, p.191, the purpose of this function is to "determine the method combination object used by a generic function". In practice however (at least in SBCL), it's not doing that either (see below), and anyway, determining the method combination object used by a generic function is better achieved by using the GENERIC-FUNCTION-METHOD-COMBINATION accessor.

So this protocol doesn't seem to make any sense. In order to understand what to make of all this, I looked at how SBCL handles method combinations (I don't know what other vendors do) and here is what I found. When you call DEFINE-METHOD-COMBINATION for some NAME, SBCL creates a new method for FIND-METHOD-COMBINATION, eql-specialized on that NAME. This method is encapsulated in a closure containing the method combination's definition, ignores its first argument (the generic function meta-object; that's why I said that it's not really doing what the AMOP says it does), recreates and returns a new method combination object on the fly every time it is called.

This has several consequences. First, the notion of "method combination named NAME" doesn't really make sense. Method combinations don't have a proper namespace. Every time you create a new generic function, the method combination essentially becomes local to that function. Redefining a method combination has no effect on existing generic functions using a method combination of the same NAME. To put it differently, you can end up with several generic functions using different yet equally named method combinations.

In Declt, I'm now assuming that programmers behave and only define method combinations once. Which brings us to the second consequence. With that assumption in mind, the "proper" way to introspect a NAME for a method combination definition (at least in SBCL) is to first look for the existence of a FIND-METHOD-COMBINATION method eql-specialized on that NAME, and then call it to retrieve the actual definition.

I haven't thought about it a lot yet, but it would be interesting to investigate the idea of cleaning this up a bit. I mean, establishing a real global namespace for method combinations, creating an alternate FIND-METHOD-COMBINATION protocol more along the lines of FIND-CLASS. And then, it could also be interesting to investigate on the potential applications of allowing method combination redefinition to affect live generic functions, just like class redefinition affects live instances...

Wednesday, June 26 2013

Nice little trick du jour

This morning, I came up with a nice little trick which made my day. Perhaps this is already well known or even idiomatic, but in case it is not, it goes like this.

Sometimes, I have to read a whole Lisp file containing more than just one toplevel form and get the contents as an unprocessed list of expressions. This happens for instance when loading a DSL program that needs post-treatment.

Usually, I end up doing something like this:

(with-open-file (stream filename)
  (loop :for expr = (read stream nil :eof)
	  :if (eq expr :eof)
	    :return exprs
	  :else
	    :collect expr :into exprs)))

But for some reason, this has always felt clunky to me. I mean, instead of ogling the family jewels of this poor shy file, I should be able to read it as a whole (hint: read-delimited-list) and preserve its intimacy.

And then I realized that the only thing that prevents me from using read-delimited-list is the lack of an ending parenthesis. But as always, Common Lisp immediately comes to the rescue, this time with its rich streams API. The idea is that you can create a string stream that just contains the closing parenthesis, and concatenate this stream at the end of the file one. I can now do something like this instead:

(with-open-file (stream filename)
  (read-delimited-list #\) (make-concatenated-stream stream (make-string-input-stream ")"))))

Not only it is shorter, it also looks much more elegant to me.

Tuesday, October 23 2012

Declt 1.0b15 "Kyoto" is out

This is Declt 1.0b15, the "Kyoto" release... Declt is a reference manual generator for Common Lisp libraries.

This version underwent a major internals overhaul, required by some of the new features described below:

  • Packages sections now advertise all definitions instead of just the symbols naming them. They also advertise their use-list and used-by-list, with cross-references.
  • Conditions, structures and classes now advertise their sub- and super-classes, direct methods, initargs and slots, with cross-references.
  • Slots documentation include docstring, type, initargs, initforms, readers and writers with cross-references.
  • Declt now documents symbol macros and compiler macros.
  • The *LINK-FILES* special is gone (M-x all-hail-purely-functional-style).
  • All ASDF components now advertise their descriptions and long descriptions, if any.
  • Docstrings are displayed in a more reader-friendly fashion.
  • Documentation entries for methods are nested within the corresponding generic function entry.

Grab it at the usual place.

Wednesday, September 26 2012

Clon 1.0b23 is out

A new version of Clon, the Command-Line Options Nuker is out.

Amongst other things, the following improvements have been made:

  • Support for ABCL has been updated, now that it provides a full MOP.
  • A workaround for SBCL's CC environment variable problem has been implemented. If the variable is not set (required for sb-grovel), Clon now switches to restricted mode instead of aborting.
  • The DUMP macro has been extend to accept a &rest argument that will be passed on to the underlying, implementation-specific, dumping facility.
  • A contrib directory has been added, for storing unapplied patches, suggestions etc.

Grab it at the usual place.

Tuesday, September 25 2012

Declt 1.0b14 is out

I've just released a new version of Declt, my reference manual generator for ASDF systems.

This release containts some improvements based on Sabra Crolleton's feedback. The most notable improvements are support for two new license types (MIT and LGPL), a new :DECLT-NOTICE keyword argument that gives you control on the "automatically generated by Declt" notice that appears in the reference manuals, and a bug fix (missing support for empty lists in lambda-lists).

Grab it at the usual place.

- page 1 of 2

French Flag English Flag
Copyright (C) 2008 -- 2018 Didier Verna didier@lrde.epita.fr