OpenProject is the leading open source project management software.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 
openproject/app/models/query.rb

639 lines
24 KiB

#-- encoding: UTF-8
#-- copyright
# OpenProject is a project management system.
# Copyright (C) 2012-2015 the OpenProject Foundation (OPF)
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License version 3.
#
# OpenProject is a fork of ChiliProject, which is a fork of Redmine. The copyright follows:
# Copyright (C) 2006-2013 Jean-Philippe Lang
# Copyright (C) 2010-2013 the ChiliProject Team
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; either version 2
# of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
#
# See doc/COPYRIGHT.rdoc for more details.
#++
class Query < ActiveRecord::Base
include Queries::WorkPackages::AvailableFilterOptions
# referenced in plugin patches - currently there are only work package queries and filters
alias_method :available_filters, :available_work_package_filters
@@user_filters = %w{assigned_to_id author_id watcher_id responsible_id}.freeze
belongs_to :project
belongs_to :user
has_one :query_menu_item, -> { order('name') },
class_name: 'MenuItems::QueryMenuItem',
dependent: :delete, foreign_key: 'navigatable_id'
serialize :filters, Queries::WorkPackages::FilterSerializer
serialize :column_names
serialize :sort_criteria, Array
validates :name, presence: true
validates_length_of :name, maximum: 255
validate :validate_work_package_filters
# WARNING: sortable should not contain a column called id (except for the
# work_packages.id column). Otherwise naming collisions can happen when AR
# optimizes a query into two separate DB queries (e.g. when joining tables).
# The first query is employed to fetch all ids which are used as filters in
# the second query. The columns mentioned in sortable are used for the first
# query. If such a statement selects from two coluns named <table name>.id
# the second one is taken to get the ids of the work packages.
@@available_columns = [
QueryColumn.new(:id,
sortable: "#{WorkPackage.table_name}.id",
groupable: false),
QueryColumn.new(:project,
sortable: "#{Project.table_name}.name",
groupable: true),
QueryColumn.new(:subject,
sortable: "#{WorkPackage.table_name}.subject"),
QueryColumn.new(:type,
sortable: "#{::Type.table_name}.position",
groupable: true),
QueryColumn.new(:parent,
sortable: ["#{WorkPackage.table_name}.root_id",
"#{WorkPackage.table_name}.lft ASC"],
default_order: 'desc'),
QueryColumn.new(:status,
sortable: "#{Status.table_name}.position",
groupable: true),
QueryColumn.new(:priority,
sortable: "#{IssuePriority.table_name}.position",
default_order: 'desc',
groupable: true),
QueryColumn.new(:author,
sortable: ["#{User.table_name}.lastname",
"#{User.table_name}.firstname",
"#{WorkPackage.table_name}.author_id"],
groupable: true),
QueryColumn.new(:assigned_to,
sortable: ["#{User.table_name}.lastname",
"#{User.table_name}.firstname",
"#{WorkPackage.table_name}.assigned_to_id"],
groupable: true),
QueryColumn.new(:responsible,
sortable: ["#{User.table_name}.lastname",
"#{User.table_name}.firstname",
"#{WorkPackage.table_name}.responsible_id"],
groupable: true,
join: 'LEFT OUTER JOIN users as responsible ON ' +
"(#{WorkPackage.table_name}.responsible_id = responsible.id)"),
QueryColumn.new(:updated_at,
sortable: "#{WorkPackage.table_name}.updated_at",
default_order: 'desc'),
QueryColumn.new(:category,
sortable: "#{Category.table_name}.name",
groupable: true),
QueryColumn.new(:fixed_version,
sortable: ["#{Version.table_name}.effective_date",
"#{Version.table_name}.name"],
default_order: 'desc',
groupable: true),
# Put empty start_dates in the far future rather than in the far past
QueryColumn.new(:start_date,
# Put empty start_dates in the far future rather than in the far past
sortable: ["CASE WHEN #{WorkPackage.table_name}.start_date IS NULL
THEN 1
ELSE 0 END",
"#{WorkPackage.table_name}.start_date"]),
QueryColumn.new(:due_date,
# Put empty due_dates in the far future rather than in the far past
sortable: ["CASE WHEN #{WorkPackage.table_name}.due_date IS NULL
THEN 1
ELSE 0 END",
"#{WorkPackage.table_name}.due_date"]),
QueryColumn.new(:estimated_hours,
sortable: "#{WorkPackage.table_name}.estimated_hours",
summable: true),
QueryColumn.new(:spent_hours,
sortable: false,
summable: false),
QueryColumn.new(:done_ratio,
sortable: "#{WorkPackage.table_name}.done_ratio",
groupable: true),
QueryColumn.new(:created_at,
sortable: "#{WorkPackage.table_name}.created_at",
default_order: 'desc'),
]
cattr_reader :available_columns
def initialize(attributes = nil, options = {})
super
add_default_filter if options[:initialize_with_default_filter]
end
def add_default_filter
self.filters = [Queries::WorkPackages::Filter.new('status_id', operator: 'o', values: [''])] if filters.blank?
end
def validate_work_package_filters
filters.each do |filter|
unless filter.valid?
messages = filter.errors.messages.values.flatten.join(" #{I18n.t('support.array.sentence_connector')} ")
cf_id = custom_field_id filter
if cf_id && CustomField.find(cf_id)
attribute_name = CustomField.find(cf_id).name
errors.add :base, attribute_name + I18n.t(default: ' %{message}', message: messages)
else
attribute_name = WorkPackage.human_attribute_name(filter.field)
errors.add :base, errors.full_message(attribute_name, messages)
end
end
end
end
def editable_by?(user)
return false unless user
# Admin can edit them all and regular users can edit their private queries
return true if user.admin? || (!is_public && user_id == user.id)
# Members can not edit public queries that are for all project (only admin is allowed to)
is_public && !for_all? && user.allowed_to?(:manage_public_queries, project)
end
def add_filter(field, operator, values)
return unless work_package_filter_available?(field)
if filter = filter_for(field)
filter.operator = operator
filter.values = values
else
filters << Queries::WorkPackages::Filter.new(field, operator: operator, values: values)
end
end
def add_short_filter(field, expression)
return unless expression
parms = expression.scan(/\A(o|c|!\*|!|\*)?(.*)\z/).first
add_filter field, (parms[0] || '='), [parms[1] || '']
end
# Add multiple filters using +add_filter+
def add_filters(fields, operators, values)
values ||= {}
if fields.is_a?(Array) && operators.is_a?(Hash) && values.is_a?(Hash)
fields.each do |field|
add_filter(field, operators[field], values[field])
end
end
end
def has_filter?(field)
filters.present? && filters.any? { |filter| filter.field.to_s == field.to_s }
end
def filter_for(field)
(filters || []).detect { |filter| filter.field.to_s == field.to_s }
end
# Deprecated
def operator_for(field)
warn '#operator_for is deprecated. Query the filter object directly, instead.'
filter_for(field).try :operator
end
# Deprecated
def values_for(field)
warn '#values_for is deprecated. Query the filter object directly, instead.'
filter_for(field).try :values
end
def label_for(field)
label = available_work_package_filters[field][:name] if work_package_filter_available?(field)
label ||= field.gsub(/\_id\z/, '')
end
def normalized_name
name.parameterize.underscore
end
def available_columns
return @available_columns if @available_columns
@available_columns = ::Query.available_columns
@available_columns += if project
project.all_work_package_custom_fields
else
WorkPackageCustomField.all
end.map { |cf| ::QueryCustomFieldColumn.new(cf) }
# have to use this instead of
# #select! as #select! can return nil
@available_columns = @available_columns.select(&:available?)
end
def self.available_columns=(v)
self.available_columns = (v)
end
def self.add_available_column(column)
available_columns << (column) if column.is_a?(QueryColumn)
end
# Returns an array of columns that can be used to group the results
def groupable_columns
available_columns.select(&:groupable)
end
# Returns a Hash of columns and the key for sorting
def sortable_columns
{ 'id' => "#{WorkPackage.table_name}.id" }.merge(available_columns.inject({}) {|h, column|
h[column.name.to_s] = column.sortable
h
})
end
def columns
if has_default_columns?
available_columns.select do |c|
# Adds the project column by default for cross-project lists
Setting.work_package_list_default_columns.include?(c.name.to_s) || (c.name == :project && project.nil?)
end
else
# preserve the column_names order
column_names.map { |name| available_columns.find { |col| col.name == name } }.compact
end
end
def column_names=(names)
if names.present?
names = names.inject([]) { |out, e| out += e.to_s.split(',') }
names = names.select { |n| n.is_a?(Symbol) || !n.blank? }
names = names.map { |n| n.is_a?(Symbol) ? n : n.to_sym }
# Set column_names to nil if default columns
if names.map(&:to_s) == Setting.work_package_list_default_columns
names = nil
end
end
write_attribute(:column_names, names)
end
def has_column?(column)
column_names && column_names.include?(column.name)
end
def has_default_columns?
column_names.nil? || column_names.empty?
end
##
# Returns the columns involved in this query, including those only needed for sorting or grouping
# puposes, not only the ones displayed (see :columns).
def involved_columns
columns = self.columns.map(&:name)
columns << group_by.to_sym if group_by
columns += sort_criteria.map { |x| x.first.to_sym }
columns.uniq
end
def sort_criteria=(arg)
c = []
if arg.is_a?(Hash)
arg = arg.keys.sort.map { |k| arg[k] }
end
c = arg.select { |k, _o| !k.to_s.blank? }.slice(0, 3).map { |k, o| [k.to_s, o == 'desc' ? o : 'asc'] }
write_attribute(:sort_criteria, c)
end
def sort_criteria
read_attribute(:sort_criteria) || []
end
def sort_criteria_sql
criteria = SortHelper::SortCriteria.new
criteria.available_criteria = sortable_columns
criteria.criteria = sort_criteria
criteria.to_sql
end
def sort_criteria_key(arg)
sort_criteria && sort_criteria[arg] && sort_criteria[arg].first
end
def sort_criteria_order(arg)
sort_criteria && sort_criteria[arg] && sort_criteria[arg].last
end
# Returns the SQL sort order that should be prepended for grouping
def group_by_sort_order
if grouped? && (column = group_by_column)
Array(column.sortable).map { |s| "#{s} #{column.default_order}" }.join(',')
end
end
# Returns true if the query is a grouped query
def grouped?
!group_by_column.nil?
end
def display_sums?
display_sums && any_summable_columns?
end
def any_summable_columns?
Setting.work_package_list_summable_columns.any?
end
def group_by_column
groupable_columns.detect { |c| c.groupable && c.name.to_s == group_by }
end
def group_by_statement
group_by_column.try(:groupable)
end
def project_statement
project_clauses = []
subproject_filter = filter_for 'subproject_id'
if project && !project.descendants.active.empty?
ids = [project.id]
if subproject_filter
case subproject_filter.operator
when '='
# include the selected subprojects
ids += subproject_filter.values.each(&:to_i)
when '!*'
# main project only
else
# all subprojects
ids += project.descendants.pluck(:id)
end
elsif Setting.display_subprojects_work_packages?
ids += project.descendants.pluck(:id)
end
project_clauses << "#{Project.table_name}.id IN (%s)" % ids.join(',')
elsif project
project_clauses << "#{Project.table_name}.id = %d" % project.id
end
project_clauses << WorkPackage.visible_condition(User.current,
project: project,
with_subprojects: !subproject_filter.nil?)
project_clauses.join(' AND ')
end
def statement
# filters clauses
filters_clauses = []
filters.each do |filter|
field = filter.field.to_s
next if field == 'subproject_id'
operator = filter.operator
values = filter.values ? filter.values.clone : [''] # HACK - some operators don't require values, but they are needed for building the statement
# "me" value substitution
if @@user_filters.include? field
if values.delete('me')
if User.current.logged?
values.push(User.current.id.to_s)
values += User.current.group_ids.map(&:to_s) if field == 'assigned_to_id'
else
values.push('0')
end
end
end
sql = ''
if field =~ /\Acf_(\d+)\z/
# custom field
db_table = CustomValue.table_name
db_field = 'value'
is_custom_filter = true
sql << "#{WorkPackage.table_name}.id IN (SELECT #{WorkPackage.table_name}.id FROM #{WorkPackage.table_name} LEFT OUTER JOIN #{db_table} ON #{db_table}.customized_type='WorkPackage' AND #{db_table}.customized_id=#{WorkPackage.table_name}.id AND #{db_table}.custom_field_id=#{$1} WHERE "
sql << sql_for_field(field, operator, values, db_table, db_field, true) + ')'
elsif field == 'watcher_id'
db_table = Watcher.table_name
db_field = 'user_id'
if User.current.admin?
# Admins can always see all watchers
sql << "#{WorkPackage.table_name}.id #{operator == '=' ? 'IN' : 'NOT IN'} (SELECT #{db_table}.watchable_id FROM #{db_table} WHERE #{db_table}.watchable_type='WorkPackage' AND #{sql_for_field field, '=', values, db_table, db_field})"
else
sql_parts = []
if User.current.logged? && user_id = values.delete(User.current.id.to_s)
# a user can always see his own watched issues
sql_parts << "#{WorkPackage.table_name}.id #{operator == '=' ? 'IN' : 'NOT IN'} (SELECT #{db_table}.watchable_id FROM #{db_table} WHERE #{db_table}.watchable_type='WorkPackage' AND #{sql_for_field field, '=', [user_id], db_table, db_field})"
end
# filter watchers only in projects the user has the permission to view watchers in
sql_parts << <<-SQL
#{WorkPackage.table_name}.id #{operator == '=' ? 'IN' : 'NOT IN'}
(SELECT #{db_table}.watchable_id
FROM #{db_table}
WHERE #{db_table}.watchable_type='WorkPackage'
AND #{sql_for_field field, '=', values, db_table, db_field})
AND #{Project.table_name}.id IN
(#{Project.allowed_to_condition(User.current, :view_work_package_watchers)})
SQL
sql << "(#{sql_parts.join(' OR ')})"
end
elsif field == 'member_of_group' # named field
if operator == '*' # Any group
groups = Group.all
operator = '=' # Override the operator since we want to find by assigned_to
elsif operator == '!*'
groups = Group.all
operator = '!' # Override the operator since we want to find by assigned_to
else
groups = Group.where(id: values)
end
groups ||= []
members_of_groups = groups.inject([]) {|user_ids, group|
if group && group.user_ids.present?
user_ids << group.user_ids
end
user_ids.flatten.uniq.compact
}.sort.map(&:to_s)
sql << '(' + sql_for_field('assigned_to_id', operator, members_of_groups, WorkPackage.table_name, 'assigned_to_id', false) + ')'
elsif field == 'assigned_to_role' # named field
roles = Role.givable
if operator == '*' # Any Role
operator = '=' # Override the operator since we want to find by assigned_to
elsif operator == '!*' # No role
operator = '!' # Override the operator since we want to find by assigned_to
else
roles = roles.where(id: values)
end
roles ||= []
members_of_roles = roles.inject([]) {|user_ids, role|
if role && role.members
user_ids << if project_id
role.members.reject { |m| m.project_id != project_id }.map(&:user_id)
else
role.members.map(&:user_id)
end
end
user_ids.flatten.uniq.compact
}.sort.map(&:to_s)
sql << '(' + sql_for_field('assigned_to_id', operator, members_of_roles, WorkPackage.table_name, 'assigned_to_id', false) + ')'
else
# regular field
db_table = WorkPackage.table_name
db_field = field
sql << '(' + sql_for_field(field, operator, values, db_table, db_field) + ')'
end
filters_clauses << sql
end if filters.present? and valid?
(filters_clauses << project_statement).join(' AND ')
end
# Returns the result set
# Valid options are :order, :include, :conditions
def results(options = {})
Results.new(self, options)
end
# Returns the journals
# Valid options are :order, :offset, :limit
def work_package_journals(options = {})
query = Journal.includes(:user)
.where(journable_type: WorkPackage.to_s)
.joins('INNER JOIN work_packages ON work_packages.id = journals.journable_id')
.joins('INNER JOIN projects ON work_packages.project_id = projects.id')
.joins('INNER JOIN users AS authors ON work_packages.author_id = authors.id')
.joins('INNER JOIN types ON work_packages.type_id = types.id')
.joins('INNER JOIN statuses ON work_packages.status_id = statuses.id')
.where(statement)
.order(options[:order])
.limit(options[:limit])
.offset(options[:offset])
.references(:users)
query
rescue ::ActiveRecord::StatementInvalid => e
raise ::Query::StatementInvalid.new(e.message)
end
# Note: Convenience method to allow the angular front end to deal with query menu items in a non implementation-specific way
def starred
!!query_menu_item
end
private
def custom_field_id(filter)
matchdata = /cf\_(?<id>\d+)/.match(filter.field.to_s)
matchdata.nil? ? nil : matchdata[:id]
end
# Helper method to generate the WHERE sql for a +field+, +operator+ and a +values+ array
def sql_for_field(field, operator, values, db_table, db_field, is_custom_filter = false)
# code expects strings (e.g. for quoting), but ints would work as well: unify them here
values = values.map(&:to_s)
sql = ''
case operator
when '='
if values.present?
if values.include?('-1')
sql = "#{db_table}.#{db_field} IS NULL OR "
end
sql += "#{db_table}.#{db_field} IN (" + values.map { |val| "'#{connection.quote_string(val)}'" }.join(',') + ')'
else
# empty set of allowed values produces no result
sql = '0=1'
end
when '!'
if values.present?
sql = "(#{db_table}.#{db_field} IS NULL OR #{db_table}.#{db_field} NOT IN (" + values.map { |val| "'#{connection.quote_string(val)}'" }.join(',') + '))'
else
# empty set of forbidden values allows all results
sql = '1=1'
end
when '!*'
sql = "#{db_table}.#{db_field} IS NULL"
sql << " OR #{db_table}.#{db_field} = ''" if is_custom_filter
when '*'
sql = "#{db_table}.#{db_field} IS NOT NULL"
sql << " AND #{db_table}.#{db_field} <> ''" if is_custom_filter
when '>='
if is_custom_filter
sql = "#{db_table}.#{db_field} != '' AND CAST(#{db_table}.#{db_field} AS decimal(60,4)) >= #{values.first.to_f}"
else
sql = "#{db_table}.#{db_field} >= #{values.first.to_f}"
end
when '<='
if is_custom_filter
sql = "#{db_table}.#{db_field} != '' AND CAST(#{db_table}.#{db_field} AS decimal(60,4)) <= #{values.first.to_f}"
else
sql = "#{db_table}.#{db_field} <= #{values.first.to_f}"
end
when 'o'
sql = "#{Status.table_name}.is_closed=#{connection.quoted_false}" if field == 'status_id'
when 'c'
sql = "#{Status.table_name}.is_closed=#{connection.quoted_true}" if field == 'status_id'
when '>t-'
sql = date_range_clause(db_table, db_field, - values.first.to_i, 0)
when '<t-'
sql = date_range_clause(db_table, db_field, nil, - values.first.to_i)
when 't-'
sql = date_range_clause(db_table, db_field, - values.first.to_i, - values.first.to_i)
when '>t+'
sql = date_range_clause(db_table, db_field, values.first.to_i, nil)
when '<t+'
sql = date_range_clause(db_table, db_field, 0, values.first.to_i)
when 't+'
sql = date_range_clause(db_table, db_field, values.first.to_i, values.first.to_i)
when 't'
sql = date_range_clause(db_table, db_field, 0, 0)
when 'w'
from = l(:general_first_day_of_week) == '7' ?
# week starts on sunday
((Date.today.cwday == 7) ? Time.now.at_beginning_of_day : Time.now.at_beginning_of_week - 1.day) :
# week starts on monday (Rails default)
Time.now.at_beginning_of_week
sql = "#{db_table}.#{db_field} BETWEEN '%s' AND '%s'" % [connection.quoted_date(from), connection.quoted_date(from + 7.days)]
when '~'
sql = "LOWER(#{db_table}.#{db_field}) LIKE '%#{connection.quote_string(values.first.to_s.downcase)}%'"
when '!~'
sql = "LOWER(#{db_table}.#{db_field}) NOT LIKE '%#{connection.quote_string(values.first.to_s.downcase)}%'"
end
sql
end
# Returns a SQL clause for a date or datetime field.
def date_range_clause(table, field, from, to)
s = []
if from
s << ("#{table}.#{field} > '%s'" % [connection.quoted_date((Date.yesterday + from).to_time.end_of_day)])
end
if to
s << ("#{table}.#{field} <= '%s'" % [connection.quoted_date((Date.today + to).to_time.end_of_day)])
end
s.join(' AND ')
end
def connection
self.class.connection
end
def for_all?
@for_all ||= project.nil?
end
end