postmodern-devel

postmodern-devel@common-lisp.net

252 discussions

Fwd: How to connect cl+ssl with host ca verification?
by Tim Hawes 28 Mar '22

28 Mar '22

Sorry Sabra, I didn't intend to reply to you directly. ---------- Forwarded message --------- From: Tim Hawes <trhawes(a)gmail.com> Date: Mon, Mar 21, 2022 at 9:35 PM Subject: Re: How to connect cl+ssl with host ca verification? To: Sabra Crolleton <sabra.crolleton(a)gmail.com> That works, but it looks like the root ca has to be added to the system. If using psql I could specify where all the certs I want to use to connect with like so: $>psql "port=5432 host=localhost user=postgres sslcert=./test/client.crt sslkey=./test/client.key sslrootcert=./test/server.crt sslmode=verify-full" Is there an equivalent in Postmodern? On Mon, Mar 21, 2022 at 8:36 PM Sabra Crolleton <sabra.crolleton(a)gmail.com> wrote: > The cl-postgres (open-database ...) function has :use-ssl as a key > parameter where the parameter values: > > - :try means if the server supports it > - :require means use provided ssl certificate with no verification > - :yes means verify that the server cert is issued by a trusted CA, > but does not verify the server hostname > - :full means expect a CA-signed cert for the supplied hostname and > verify the server hostname > > This is mirrored in the postmodern (connect ...) function. > > If I understand your question correctly, you want to use :full as the > parameter passed to :use-ssl. E.g. > >> (connect "test_db" "test-user" "test-password" "192.168.5.223" >> :port 5434 :pooled-p t :use-ssl :full) >> >> But maybe I am not understanding your question correctly. > > Sabra > > On Mon, Mar 21, 2022 at 7:50 PM Tim Hawes <trhawes(a)gmail.com> wrote: > >> I am not finding any helpful information in how to connect to PostgreSQL >> using client key/client cert and a root ca for verifying the host with >> Postmodern. Can someone send me an example? >> >

2 5

How to connect cl+ssl with host ca verification?
by Tim Hawes 21 Mar '22

21 Mar '22

I am not finding any helpful information in how to connect to PostgreSQL using client key/client cert and a root ca for verifying the host with Postmodern. Can someone send me an example?

1 0

Postmodern Query Macro and binary parameters
by Sabra Crolleton 07 Jun '21

07 Jun '21

There has been a long standing issue about not having the ability to pass parameters to Postgresql in binary format. Currently all parameters are passed as text and Postgresql figures out what type they should be. Cyrus raised the question about whether passing parameters as binary rather than text should be set globally or on a query by query basis. He also raises the question of whether the QUERY macro could be more efficiently re-written. Your thoughts are requested. Sabra

1 0

loading problem
by Jans Aasman 30 Nov '20

30 Nov '20

Hi: this is my first time posting here so I have no idea whom to address this to. Quicklisp was my friend in downloading and installing postmodern in a our case sensitive ACL 10.1 SMP The only problem I ran into was that in the file normalize-backend.lisp there was one upper case DO and one upper case WHEN. After fixing that I could load the whole thing. Question: could this be fixed in your source somewhere? Cheers, Jans [package uax-15]..; While file-compiling #'compose-hangul in #P"/home/test/quicklisp/dists/quicklisp/software/uax-15-20200325-git/src/normalize-backend.lisp" ; starting at file character position 563: Error: DO is an unknown keyword in LOOP macro. Current LOOP context: DO (tagbody (when (<= 0 l-index (1- l-count)) (let ((v-index (- ch v-base))) (when (<= 0 v-index (1- v-count)) (setf last (+ s-base (* (+ (* l-index v-count) v-index) t-count))) (setf (car new-chars) last) (go :end)))) (when (and (<= 0 s-index (1- s-count)) (zerop (mod s-index t-count))) (let ((t-index (- ch t-base))) (when (< 0 t-index t-count) (setf last (+ last t-index)) (setf (car new-chars) last) (go :end)))) (push (setf last ch) new-chars) :end). [condition type: program-error] .; While file-compiling #'compose in #P"/home/test/quicklisp/dists/quicklisp/software/uax-15-20200325-git/src/normalize-backend.lisp" ; starting at file character position 3336: Error: WHEN is an unknown keyword in LOOP macro. Current LOOP context: WHEN ch-left. [condition type: program-error] Restart actions (select using :continue): 0: retry the compilation of /home/test/quicklisp/dists/quicklisp/software/uax-15-20200325-git/src/normalize-backend.lisp 1: continue compiling /home/test/quicklisp/dists/quicklisp/software/uax-15-20200325-git/src/normalize-backend.lisp but generate no output file 2: Retry compiling #<cl-source-file "uax-15" "src" "normalize-backend">. 3: Continue, treating compiling #<cl-source-file "uax-15" "src" "normalize-backend"> as having been successful. 4: Retry ASDF operation. 5: Retry ASDF operation after resetting the configuration. 6: Give up on "postmodern" 7: Return to Top Level (an "abort" restart). 8: Abort entirely from this (lisp) process. [1] uax-15(7): :cont 0 ............................... [package usocket]................................. [package cl-postgres].............................

1 0

How to create dao class with identity column as primary key?
by Timo Myyrä 11 Jun '20

11 Jun '20

Hi, I'm learning to use postmodern as part of my hobby project but I've hit a small bump in the process. I can have the identity column defined for table without dao with something like: (s-sql:sql (:create-table (:if-not-exists 'user) ((id :type int :identity-always t) (username :type text) (:primary-key id))) But I intent to use dao classes so it would be nice to have identity column specified as part to defclass slot options. Is there some way to create dao class with identity column? Br, Timo M

2 5

New date/time integration for cl-postgres
by Chaitanya Gupta 19 Feb '19

19 Feb '19

Hi, I created a new date/time integration for cl-postgres. https://github.com/chaitanyagupta/cl-postgres-datetime There's nothing fancy -- it just uses local-time to parse timestamptz and simple-date for other types (which don't use a timezone). Why? Because neither local-time nor simple-date alone work well with all the date/time types that Postgres supports. local-time's TIMESTAMP is a natural fit for timestamptz, and while non-timezone types like timestamp and date could be parsed into it, the user has to remember to always use +UTC-ZONE+ when decoding such a timestamp (the values of decoded components may mismatch with the original if the default time zone were different from UTC). simple-date has no concept of time zones, so while it works well for every type that doesn't need a timezone, it fails badly when it comes to timestamptz. Hope someone finds it useful. Chaitanya --- lisper.in deftask.com

2 5

Does anyone use the hstore operators in s-sql?
by Sabra Crolleton 23 Oct '18

23 Oct '18

I am looking at the hstore operators in s-sql. Does anyone use them? If so, could you send examples of how you are using them? I want to make sure that if I touch those operators, I do not accidentally break anyone's code. Sabra

2 2

:create-table extension
by Steven Nunez 12 Feb '18

12 Feb '18

Greetings all, I am updating the :create-table macro to work with Greenplum's CREATE TABLE syntax. For the most part, this is identical except for two important table attributes: DISTRIBUTED BY (column, [ ... ] ) | DISTRIBUTED RANDOMLY These tell Greenplum how to distribute the table across the cluster, and getting these right is critical for good performance. Here is a SQL example using the FAA dataset: create table faa.d_airports ( AirportID integer, Name text, City text, Country text, airport_code text, ICOA_code text, Latitude float8, Longitude float8, Altitude float8, TimeZoneOffset float, DST_Flag text , TZ text ) distributed by (airport_code); The challenge is in emitting the 'distributed by (...)' or 'distributed randomly' after the closing parenthesis in the SQL. The S-SQL macro, with the correct location is: (def-sql-op :create-table (name (&rest columns) &rest options) (when (null columns) (sql-error "No columns defined for table ~A." name)) `("CREATE TABLE " ,(to-sql-name name) " (" ,@(loop :for ((column-name . args) . rest) :on columns :append (expand-table-column column-name args) :if rest :collect ", ") ,@(loop :for ((option . args)) :on options :collect ", " :append (expand-table-constraint option args)) ")" ; DISTRIBUTED BY needs to go here )) Initially I thought I could modify expand-table-constraints in a way that would do the trick, but I could not find a way to do this without either hardcoding these two cases or modifying the function signature. Modifying the signature of the :create-table macro seems the best choice, and then add another invocation of expand-table-constraint after the closing parentheses, but that is a breaking change. Alternatively we could have a :create-gp-table, but I would rather not diverge. So, I am here to ask: am I missing something obvious? If not, what is the consensus for moving forward? I supposed I would be most in favor of modifying the signature for :create-table. Regards, Steve P.S. There is also an open issue for create table that might have a need for something similar. I do not know, but it is worth considering both of them whilst making the change.

1 0

test for the mailing list
by Sabra Crolleton 11 Feb '18

11 Feb '18

This is just a test.

1 0

Supporting ~/.pgpass?
by Johann 'Myrkraverk' Oskarsson 02 Sep '15

02 Sep '15

Hi all, I've previously blogged one naive parser for ~/.pgpass, at http://www.myrkraverk.com/blog/2015/03/postgresqls-pgpass-parser-in-common-… Is it possible to incorporate that, or a better parser, into Postmodern? And thank you, those who answered my question about (:as 'tbl 't), -- Johann I'm not from the internet, I just work there.

2 3

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

postmodern-devel