1
nginx
      openssl
      Fri, 04 Aug 2017 13:48:57 GMT
      
        自定义编译nginx时，出现了如下报错
1
2
3
relocation R_X86_64_32 against `.rodata' can not be used when making a shared object; recompile with -fPIC
/path/to/openssl/.openssl/lib/libssl.a: error adding symbols: Bad value
collect2: error: ld returned 1 exit status
打算直接修改Makefile文件时，找到了askubuntu，新的解决方法，尝试之后，编译通过，在这里记录解决方法。
在nginx目录下的 auto/options 文件中找到这么一行
1
--with-openssl-opt=*)            OPENSSL_OPT="$value"       ;; 
修改为如下样式
1
--with-openssl-opt=*)            OPENSSL_OPT="$value -fPIC"       ;; 
可能因为版本不同，我找到的跟链接例子中的不大一样，但是可以解决编译错误。
]]>
      
    
    
      
      https://hexo.guoqiang.info/2017/07/21/xargs-with-space/
      
        多文件使用xargs操作时，经常会出现带空格的文件名，而有文件找不到的错误提示。在 xargs
      shell
      script
      Fri, 21 Jul 2017 11:51:35 GMT
      
        多文件使用xargs操作时，经常会出现带空格的文件名，而有文件找不到的错误提示。在 stackoverflow上找到两种解决方法。在此贴上。
1.使用换行符(\n)作为分隔符
1
ls *mp3 | xargs -d '\n' mplayer
2.跟find连用时
1
2
3
-0      Change xargs to expect NUL (``\0'') characters as separators,
        instead of spaces and newlines.  This is expected to be used in
        concert with the -print0 function in find(1).
1
find . -name "*.mp3" -print0 | xargs -0 mplayer
]]>
      
    
    
      
      https://hexo.guoqiang.info/2017/07/21/docker-hexo/
      
        很早想在docker上部署hexo，但是一直没(tai)时(lan)间(le)。终于抽空部署上来，问题多多，记录一下。

1.手动拉取 nodejs 镜像 docker pull node:latest，个人喜好，后面创建镜像的时候可以]]>
      
      Qiang
      docker
      hexo
      Fri, 21 Jul 2017 10:26:56 GMT
      
        很早想在docker上部署hexo，但是一直没(tai)时(lan)间(le)。终于抽空部署上来，问题多多，记录一下。
1.手动拉取 nodejs 镜像 docker pull node:latest，个人喜好，后面创建镜像的时候可以节省时间。顺便提一下，使用daocloud，拉取速度提升了1到2个数量级。
2.安装hexo，npm install -g hexo-cli，这一句就直接给我搞了个没有权限的报错。貌似是npm的一个bug，google到了解决方法如下。
1
2
3
npm config set user 0
npm config set unsafe-perm true
npm install hexo-cli -g
3.因为已经有现成的hexo工程，并没有把工程安装写在Dockerfile中，而是进入container完成工程的重新部署。Dockerfile只是安装环境和部分软件。就不传了。  
4.全自动部署，当然是需要docker-compose了。但是使用compose运行时，提示nodejs exited with code 0，而直接用Dockerfile却可以正常运行，参考这个stackoverflow得到了解决方法。在docker-compose.yml文件增加一行tty: true，可以实现后台运行。  
5.由于桌面端没装nodejs(本来就是不想装，才用的docker)，无法运行hexo命令。在虚拟机里运行的话，每次要敲一大堆的命令。但是桌面上装上的话，又背离了使用docker的初衷。所以，还是用脚本来解决吧。在/usr/bin 下面增加一个文件，内容如下。
1
2
3
#!/bin/bash
all=$@
docker exec -i hexo bash -c "cd /path/to/hexo;hexo $all"
hexo使用的container 名字是 hexo, 通过一个all来传递所有参数，是因为直接在 -c后面用 $@时，多参数只能收到第一个参数，采用了个变通的方法。使用时，直接在桌面运行 hexo param就可以了。
大功告成，用此docker完成本文。
]]>
      
    
    
      
      https://hexo.guoqiang.info/2017/02/23/git-ssl-issuer/
      
        搭建了本地gogs，看看使用情况，gogs自签证书只需要一个指令，就顺便加上了https，结果


      
      Qiang
      Thu, 23 Feb 2017 11:00:00 GMT
      
        搭建了本地gogs，看看使用情况，gogs自签证书只需要一个指令，就顺便加上了https，结果
1
git config --global http.sslVerify false
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/09/22/selenium%E5%AD%A6%E4%B9%A0%E8%AE%B0%E5%BD%95/
      
        1.鼠标悬停的实现方式 


a) 
  
1
      
      Qiang
      selenium
      python
      selenium
      python
      Thu, 22 Sep 2016 21:54:02 GMT
      
        1.鼠标悬停的实现方式 
a) 
  
1
2
3
test = browser.find_element_by_id("qmenu")
hover = ActionChains(browser).move_to_element(test)
hover.perform()
b)
1
2
3
4
5
6
from selenium.webdriver.common.action_chains import ActionChains
def hover(self):
    wd = webdriver_connection.connection
    element = wd.find_element_by_link_text(self.locator)
    hov = ActionChains(wd).move_to_element(element)
    hov.perform()
2.翻页功能
1
driver.execute_script("window.scrollTo(0, document.body.scrollHeight);")
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/05/13/gitlab%E5%8D%87%E7%BA%A7%E5%90%8Ehttp%E6%96%B9%E5%BC%8F502%E9%94%99%E8%AF%AF/
      
        gitlab从8.1升级到8.7.4，ssh方式已经可以提交获取代码了，但是通过http方式还是502。

查询nginx后台日志显示 connect() to unix://var/opt/gitlab/gitlab-git-http-server]]>
      
      Qiang
      gitlab
      Fri, 13 May 2016 14:11:32 GMT
      
        gitlab从8.1升级到8.7.4，ssh方式已经可以提交获取代码了，但是通过http方式还是502。
查询nginx后台日志显示 connect() to unix://var/opt/gitlab/gitlab-git-http-server/socket failed (13: Permission denied) 把nginx用户加到文件所在的组后，再次尝试依然是502，但是错误日志内容有了变化 connect() to unix://var/opt/gitlab/gitlab-git-http-server/socket failed (111: Connection refused) 连接拒绝了。google搜结果，在这里找到了答案。修改nginx配置，把 /var/opt/gitlab/gitlab-git-http-server/socket 替换为 /var/opt/gitlab/gitlab-workhorse/socket 就可以了。
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/04/07/linux-script/
      
        1.使用cp时强制覆盖，两种方案:(来源)
a) /bin/cp -rf /zzz/zzz/* /xxx/xxx
b) /bi]]>
      
      Qiang
      linux
      Thu, 07 Apr 2016 17:04:20 GMT
      
        1.使用cp时强制覆盖，两种方案:(来源)
a) /bin/cp -rf /zzz/zzz/* /xxx/xxx
b) /bin/cp -rf /zzz/zzz/* /xxx/xxx
原来当使用root登陆时，环境变量中会有一条 alias cp = cp -i ，用于防止管理员误操作。
2.crontab环境变量
  crontab执行命令时，并不适用当前用户的环境变量，编写任务时最好使用全路径。确实需要环境变量时，可以在crontab文件的指令中直接执行 source ~/.bash_profile 或 /etc/profile等加载用户环境变量。
3.crontab中的run-parts
  运行指定目录中的所有脚本，注意脚本文件中不能含有’.’，因为不含有任何参数的run-parts将会忽略他们。例如backup.sh这个脚本是不会被执行的。
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/03/12/nginx-413/
      
         Nginx出现的413 Request Entity Too Large错误,这个错误一般在上传文件的时候出现，打开nginx主配置文件nginx.conf，找到http{}段，添加
解决方法就是
打开nginx主配置文件nginx.conf，一般在/]]>
      
      Qiang
      nginx
      Sat, 12 Mar 2016 11:19:11 GMT
      
         Nginx出现的413 Request Entity Too Large错误,这个错误一般在上传文件的时候出现，打开nginx主配置文件nginx.conf，找到http{}段，添加
解决方法就是
打开nginx主配置文件nginx.conf，一般在/usr/local/nginx/conf/nginx.conf这个位置，找到http{}段，修改或者添加
代码如下复制代码
client_max_body_size 2m;
然后重启nginx，
代码如下复制代码
sudo /etc/init.d/nginxd reload
即可。
要是以php运行的话，这个大小client_max_body_size要和php.ini中的如下值的最大值差不多或者稍大，这样就不会因为提交数据大小不一致出现错误。
代码如下复制代码
post_max_size = 2M
upload_max_filesize = 2M
重启NGINX
代码如下复制代码
kill -HUP cat /usr/local/nginx/nginx.pid 
恢复正常
除非特别声明，PHP100新闻均为原创或投稿报道，转载请注明作者及原文链接
原文地址： http://www.php100.com/html/program/nginx/2013/0905/5516.html
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/02/29/php-parse-error/
      
        PHP模版文件打开后出现”Parse error: syntax error, unexpected end of file in ‘file path’”,在php
      php
      Mon, 29 Feb 2016 09:36:21 GMT
      
        PHP模版文件打开后出现”Parse error: syntax error, unexpected end of file in ‘file path’”,在stackoverflow找到了解决方法。  
在代码的行尾要尽量避免这种
1
{?>
和这种代码
1
<?php}
要使用空格分开，像这样
1
2
{ ?>
<?php {
还要避免使用 <? 而应该使用 <?php 。  
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/02/24/magento-directly-sql/
      
        原文DIRECT SQL QUERIES IN MAGENTO有点长，直接转过来了。
Magento’s use of data models provide a great way to access and modify data. Using aptly named methods and clever abstraction, Varien hide away the complex SQL needed to perform data operations. While this makes learning models easier, it often impacts the speed of the operation and therefore the responsiveness of your site. This is especially true when saving models that use the EAV architecture. More often that not, this cannot be avoided, however there are some situations where executing direct SQL queries would be simpler and much quicker leading to a more optimised Magento installation. An example of this is updating product prices globally in Magento. It would be easy enough to write some Magento code that looped through all products and modified the price. On a large data set, saving each individual product can take a long time and therefore make the system unusable. To combat this, it is possible to issue a direct SQL query which could update 1000’s of products in 1 or 2 seconds.
]]>
      
      Qiang
      magento
      php
      magento
      sql
      Wed, 24 Feb 2016 17:32:25 GMT
      
        原文DIRECT SQL QUERIES IN MAGENTO有点长，直接转过来了。
Magento’s use of data models provide a great way to access and modify data. Using aptly named methods and clever abstraction, Varien hide away the complex SQL needed to perform data operations. While this makes learning models easier, it often impacts the speed of the operation and therefore the responsiveness of your site. This is especially true when saving models that use the EAV architecture. More often that not, this cannot be avoided, however there are some situations where executing direct SQL queries would be simpler and much quicker leading to a more optimised Magento installation. An example of this is updating product prices globally in Magento. It would be easy enough to write some Magento code that looped through all products and modified the price. On a large data set, saving each individual product can take a long time and therefore make the system unusable. To combat this, it is possible to issue a direct SQL query which could update 1000’s of products in 1 or 2 seconds.
Database Connections In Magento
By default, Magento will automatically connect to it’s database and provide two separate resources which you can use to access data: core_read and core_write. As you can probably guess, core_read is for reading from the database while core_write is for writing to the database. It is important to ensure that you use the correct resource when reading or writing data to the database, especially when writing custom Magento extensions that will be released into the wild.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
<?php   
    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Retrieve the read connection
     */
    $readConnection = $resource->getConnection('core_read');
    
    /**
     * Retrieve the write connection
     */
    $writeConnection = $resource->getConnection('core_write');
Table names and table prefixes
When installing Magento, you are given the option to use a table prefix. A table prefix is a string of characters that is added to the start of every table name in your database. These are useful if you are installing multiple system into 1 database as it helps to distinguish each application’s data from another. Fortunately, Magento has a simple built in function which allows you to add the prefix to a given table name.
Get a table name from a string
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
<?php

    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Get the table name
     */
    $tableName = $resource->getTableName('catalog_product_entity');
    
    /**
     * if prefix was 'mage_' then the below statement
     * would print out mage_catalog_product_entity
     */
    echo $tableName;
Get a table name from an entity name
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
<?php

    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Get the table name
     */
    $tableName = $resource->getTableName('catalog/product');
    
    /**
     * if prefix was 'mage_' then the below statement
     * would print out mage_catalog_product_entity
     */
    echo $tableName;
Reading From The Database
While Magento models hide the complexity of the EAV system, they sometimes request far more data than is needed. If for example you have a product ID and want it’s SKU, it would be much quicker to run a single query to obtain this value than to load in a whole product model (the inverse of this operation is available via the product resource class).
Varien_Db_Select::fetchAll
This method takes a query as it’s parameter, executes it and then returns all of the results as an array. In the code example below, we use Varien_Db_Select::fetchAll to return all of the records in the catalog_product_entity table.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
<?php
    
    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Retrieve the read connection
     */
    $readConnection = $resource->getConnection('core_read');
    
    $query = 'SELECT * FROM ' . $resource->getTableName('catalog/product');
    
    /**
     * Execute the query and store the results in $results
     */
    $results = $readConnection->fetchAll($query);
    
    /**
     * Print out the results
     */
     var_dump($results);
Varien_Db_Select::fetchCol
This method is similar to fetchAll except that instead of returning all of the results, it returns the first column from each result row. In the code example below, we use Varien_Db_Select::fetchCol to retrieve all of the SKU’s in our database in an array.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
<?php
    /**
      * Get the resource model
      */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Retrieve the read connection
     */
    $readConnection = $resource->getConnection('core_read');
    /**
     * Retrieve our table name
     */
    $table = $resource->getTableName('catalog/product');

    /**
     * Execute the query and store the results in $results
     */
    $sku = $readConnection->fetchCol('SELECT sku FROM ' . $table . '*');
    
    /**
     * Print out the results
     */
     var_dump($results);
Try this code and look at the results. Notice how all of the SKU’s are in a single array, rather than each row having it’s own array? If you don’t understand this, try changing fetchCol for fetchAll and compare the differences.
Varien_Db_Select::fetchOne
Unlike the previous two methods, Varien_Db_Select::fetchOne returns one value from the first row only. This value is returned on it’s own and is not wrapped in an array. In the code example below, we take a product ID of 44 and return it’s SKU.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
<?php

    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Retrieve the read connection
     */
    $readConnection = $resource->getConnection('core_read');

    /**
     * Retrieve our table name
     */
    $table = $resource->getTableName('catalog/product');
    
    /**
     * Set the product ID
     */
    $productId = 44;
    
    $query = 'SELECT sku FROM ' . $table . ' WHERE entity_id = '
             . (int)$productId . ' LIMIT 1';
    
    /**
     * Execute the query and store the result in $sku
     */
    $sku = $readConnection->fetchOne($query);
    
    /**
     * Print the SKU to the screen
     */
    echo 'SKU: ' . $sku . '<br/>';
When trying out this example, ensure you change the product ID to an ID that exists in your database!
You may think that fetchOne works the same as fetchCol or fetchAll would if you only added 1 column to the SELECT query and added a ‘LIMIT 1’, however you would be wrong. The main difference with this function is that the value returned is the actual value, where as Varien_Db_Select::fetchCol and Varien_Db_Select::fetchAll would wrap the value in an array. To understand this a little, try swapping the method’s and comparing the results.
Writing To The Database
When saving a Magento model, there can be a lot of background data being saved that you weren’t even aware of. For example, saving a product model can take several seconds due to the amount of related data saves and indexing that needs to take place. This is okay if you need all the data saving, but if you only want to update the SKU of a product, this can be wasteful.
The example code below will show you how when given a product ID, you can alter the SKU. This is a trivial example but should illustrate how to execute write queries against your Magento database.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
<?php

    /**
     * Get the resource model
     */
    $resource = Mage::getSingleton('core/resource');
    
    /**
     * Retrieve the write connection
     */
    $writeConnection = $resource->getConnection('core_write');

    /**
     * Retrieve our table name
     */
    $table = $resource->getTableName('catalog/product');
    
    /**
     * Set the product ID
     */
    $productId = 44;
    
    /**
     * Set the new SKU
     * It is assumed that you are hard coding the new SKU in
     * If the input is not dynamic, consider using the
     * Varien_Db_Select object to insert data
     */
    $newSku = 'new-sku';
    
    $query = "UPDATE {$table} SET sku = '{$sku}' WHERE entity_id = "
             . (int)$productId;
    
    /**
     * Execute the query
     */
    $writeConnection->query($query);
To test this has worked, use the knowledge gained from the first part of this tutorial to write a query to extract the SKU that has just been changed.
Varien_Db_Select
The Varien_Db_Select, which has been touched on in this article is a far better option for extracting/wriiting information. Not only is it easy to use, it also provides a layered of security, which if used correctly, is impenetrable. More will be covered on Varien_Db_Select (aka Zend_Db_Select) in a future article.
Conclusion
Sometimes it is necessary to execute direct SQL queries in Magento, however, please be careful! The Magento model’s are there for a reason and provide a layer of security which you will have to manually add to your own direct SQL queries. Be sure to escape any user input and when possible, stick to the Magento model methods! If you can’t stick to the Magento models, consider using Varien_Db_Select; it won’t stop you making errors but it will add an almost impenetrable layer of security to your database queries.
As a side note, if you’re going to be querying the database directly, it would be a good idea to learn about Magento’s EAV database architecture.
DIRECT SQL QUERIES IN MAGENTO
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/02/24/ssl-certificate-transparency/
      
        转载自：为 SSL 站点启用 Certificate Transparency 功能
下载 n]]>
      
      Qiang
      ssl
      Wed, 24 Feb 2016 14:36:18 GMT
      
        转载自：为 SSL 站点启用 Certificate Transparency 功能
下载 nginx 源代码和 nginx-ct 的源代码：
1
2
3
4
wget http://nginx.org/download/nginx-1.9.6.tar.gz
wget -O nginx-ct.zip https://github.com/grahamedgecombe/nginx-ct/archive/master.zip
tar zxf nginx-1.9.6.tar.gz
unzip nginx-ct.zip
编译 nginx ( 你也可以带上你自己的参数，加上 ssl 和 ct module 就可以）
1
2
cd nginx-1.9.6/
./configure --with-http_v2_module --with-http_ssl_module --add-module=../nginx-ct-master
创建 SCT 的文件夹
1
mkdir /etc/ssl/scts/
下载证书提交工具
1
2
3
4
5
wget -O ct-submit.zip https://github.com/grahamedgecombe/ct-submit/archive/master.zip
unzip ct-submit.zip
cd ct-submit-master/
# 请确保已经安装 go 语言
go build
将证书提交到 Certificate Transparency Log 服务器，假设你的 SSL 证书在 /etc/ssl/server.crt （带证书链的）
1
2
3
4
5
6
7
8
9
sudo sh -c "./ct-submit-master ct.googleapis.com/aviator \
  </etc/ssl/server.crt \
  >/etc/ssl/scts/aviator.sct"
sudo sh -c "./ct-submit-master ct.googleapis.com/pilot \
  </etc/ssl/server.crt \
  >/etc/ssl/scts/pilot.sct"
sudo sh -c "./ct-submit-master ct.googleapis.com/rocketeer \
  </etc/ssl/server.crt \
  >/etc/ssl/scts/rocketeer.sct"
然后在 nginx 配置中添加
1
2
ssl_ct on;
ssl_ct_static_scts /etc/ssl/scts;
以上内容转载自：为 SSL 站点启用 Certificate Transparency 功能
可能因为使用的linux发行版不一样，我用的是centos，在执行nginx的configure时需要指定openssl的源码路径(openssl 1.0.2+，我用的是 openssl-1.0.2f，奇怪的版本号)，否则会有未定义函数的报错。
--with-openssl=/path/to/openssl
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/02/23/magento-xml-block-type/
      
        Magento在xml文件中block的type解释。结果也是在网上搜到的，按照 type="A/B"的格式来解释。
答案转自magento
      php
      magento
      Tue, 23 Feb 2016 22:46:38 GMT
      
        Magento在xml文件中block的type解释。结果也是在网上搜到的，按照 type="A/B"的格式来解释。
答案转自链接。 
1
<block type="page/html" name="root" output="toHtml" template="example/view.phtml"> 
A是一个模块的别名。在这个例子里，’page’是Mage_Page_block(在文件app/code/core/Mage/Page/etc/config.xml中定义)。
B是A模块中的一个class名，每个词的首字母大写。在这个例子中，html转换为Html，是Mage_Page_block_Html的别名。这个文件大概在app/code/core/Mage/Page/Block/Html.php ，因为在Magento中，类名可以直接转换为路径。
如果用model别名替换block别名结果则是Mage_Page_model。换成resource models和helpers也是一样的。自己写的模块如果包含block、models、helpers也应该在配置文件里(config/)定义。
链接
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/01/25/magento-failed-to-open-stream-soapclient-php/
      
        Warning: include(SoapClient.php): failed to open stream: No such file or directory

新部署magento时，出现上面一条错误，原因是未安装ph]]>
      
      Qiang
      magento
      php
      magento
      soapclient
      Mon, 25 Jan 2016 18:29:18 GMT
      
        Warning: include(SoapClient.php): failed to open stream: No such file or directory
新部署magento时，出现上面一条错误，原因是未安装php-soap，安装php-soap就没有错误了，要根据已安装的php版本来选择。
]]>
      
    
    
      
      https://hexo.guoqiang.info/2016/01/05/virtualbox-mount-shared-folder/
      
        通过下面的指令可以挂载目录，并指定所属用户和用户组

sudo mount -t vboxsf -o uid=1000,gid=1000 host_shared_folder /path/to/local/]]>
      
      Qiang
      virtualbox
      mount
      Tue, 05 Jan 2016 10:45:59 GMT
      
        通过下面的指令可以挂载目录，并指定所属用户和用户组
sudo mount -t vboxsf -o uid=1000,gid=1000 host_shared_folder /path/to/local/
]]>